bauer/dotfiles
1
0
Fork 0
mirror of https://github.com/dnlbauer/dotfiles.git synced 2026-09-10 13:35:30 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat: add claude config

Browse Source
This commit is contained in:
Daniel Bauer
2026-08-09 14:15:17 +02:00
parent 22af30aa1b
commit f165a04aca
27 changed files with 7124 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

184
dot_claude/executable_statusline.sh Executable file
View File

@@ -0,0 +1,184 @@
#!/bin/bash
# Read JSON input from stdin
input=$(cat)
# \u2500\u2500 helpers \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500
get_current_dir() {
local dir
dir=$(echo "$input" | jq -r '.workspace.project_dir // .cwd // ""')
# Bold white
printf '\033[1;37m%s\033[0m' "${dir##*/}"
}
get_model_effort() {
local model effort thinking
model=$(echo "$input" | jq -r '.model.display_name // ""')
effort=$(echo "$input" | jq -r '.effort.level // ""')
thinking=$(echo "$input" | jq -r '.thinking.enabled // false')
# Bold magenta model name; dim+italic tag
local colored_model
colored_model=$(printf '\033[1;35m%s\033[0m' "$model")
if [ -n "$effort" ]; then
printf '%s \033[0m[%s]' "$colored_model" "$effort"
elif [ "$thinking" = "true" ]; then
printf '%s \033[0m[thinking]' "$colored_model"
else
printf '%s' "$colored_model"
fi
}
get_context() {
local used
used=$(echo "$input" | jq -r '.context_window.used_percentage // empty')
if [ -z "$used" ]; then
echo "Ctx: -"
return
fi
# Read raw token counts
local used_tokens max_tokens
used_tokens=$(echo "$input" | jq -r '.context_window.total_input_tokens // empty')
max_tokens=$(echo "$input" | jq -r '.context_window.context_window_size // empty')
# Format a raw number as e.g. 85k or 1.2M
fmt_tokens() {
local n="$1"
if [ -z "$n" ] || [ "$n" = "null" ]; then
echo "?"
return
fi
awk -v n="$n" 'BEGIN {
if (n >= 1000000) {
v = n / 1000000
# one decimal place, strip trailing .0
s = sprintf("%.1f", v)
sub(/\.0$/, "", s)
print s "M"
} else if (n >= 1000) {
v = n / 1000
s = sprintf("%.1f", v)
sub(/\.0$/, "", s)
print s "k"
} else {
print n
}
}'
}
local used_fmt max_fmt
used_fmt=$(fmt_tokens "$used_tokens")
max_fmt=$(fmt_tokens "$max_tokens")
# Build a 20-cell wide loading bar
local bar_width=20
local filled
filled=$(printf "%.0f" "$(echo "$used $bar_width" | awk '{printf "%f", $1 * $2 / 100}')")
# Clamp to [0, bar_width]
[ "$filled" -lt 0 ] && filled=0
[ "$filled" -gt "$bar_width" ] && filled=$bar_width
local empty=$(( bar_width - filled ))
# Pick fill colour: green <=60, yellow <=80, red >80
local fill_color
if awk "BEGIN {exit !($used <= 60)}"; then
fill_color='\033[32m' # green
elif awk "BEGIN {exit !($used <= 80)}"; then
fill_color='\033[33m' # yellow
else
fill_color='\033[31m' # red
fi
local filled_bar="" empty_bar=""
local i
for (( i=0; i<filled; i++ )); do filled_bar="${filled_bar}▓"; done
for (( i=0; i<empty; i++ )); do empty_bar="${empty_bar}░"; done
# coloured fill blocks + dim-grey empty blocks | bold percentage | token counts
printf "Ctx: ${fill_color}%s\033[0m\033[2;37m%s\033[0m \033[1m%.0f%%\033[0m \033[2;37m%s/%s\033[0m" \
"$filled_bar" "$empty_bar" "$used" "$used_fmt" "$max_fmt"
}
get_reset_timer() {
# Five-hour window resets_at (unix epoch)
local resets_at now diff h m
resets_at=$(echo "$input" | jq -r '.rate_limits.five_hour.resets_at // empty')
if [ -z "$resets_at" ]; then
echo ""
return
fi
now=$(date +%s)
diff=$((resets_at - now))
if [ "$diff" -le 0 ]; then
# Dim white/grey
printf '\033[2;37mReset: now\033[0m'
return
fi
h=$((diff / 3600))
m=$(( (diff % 3600) / 60 ))
# Dim white/grey
printf '\033[2;37mReset: %dh%02dm\033[0m' "$h" "$m"
}
get_git_branch_info() {
local cwd
cwd=$(echo "$input" | jq -r '.cwd // ""')
if [ -z "$cwd" ]; then
echo ""
return
fi
# Run git in the session's cwd to avoid lock races
local branch changes
branch=$(git -C "$cwd" branch --show-current 2>/dev/null)
if [ -z "$branch" ]; then
echo ""
return
fi
# Count staged + unstaged + untracked changes
local status_output
status_output=$(git -C "$cwd" status --porcelain 2>/dev/null)
local change_count
change_count=$(echo "$status_output" | grep -c '^' 2>/dev/null || echo 0)
# grep -c returns 1 even for empty input on some platforms; guard it
if [ -z "$status_output" ]; then
change_count=0
fi
if [ "$change_count" -gt 0 ]; then
# Bold green branch, yellow change count
printf '\033[1;32m\u2387 %s\033[0m \033[33m(%d)\033[0m' "$branch" "$change_count"
else
# Bold green branch
printf '\033[1;32m\u2387 %s\033[0m' "$branch"
fi
}
# \u2500\u2500 assemble \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500
CURRENT_DIR=$(get_current_dir)
MODEL_EFFORT=$(get_model_effort)
CONTEXT=$(get_context)
RESET=$(get_reset_timer)
GIT_INFO=$(get_git_branch_info)
parts=()
[ -n "$CURRENT_DIR" ] && parts+=("$CURRENT_DIR")
[ -n "$MODEL_EFFORT" ] && parts+=("$MODEL_EFFORT")
[ -n "$CONTEXT" ] && parts+=("$CONTEXT")
[ -n "$RESET" ] && parts+=("$RESET")
[ -n "$GIT_INFO" ] && parts+=("$GIT_INFO")
# Join with " | "
result=""
for part in "${parts[@]}"; do
if [ -z "$result" ]; then
result="$part"
else
result="$result\033[2m | \033[0m$part"
fi
done
printf '%b\n' "$result"

21
dot_claude/settings.json Normal file
View File

@@ -0,0 +1,21 @@
{
"$schema": "https://json.schemastore.org/claude-code-settings.json",
"includeCoAuthoredBy": false,
"statusLine": {
"type": "command",
"command": "~/.claude/statusline.sh",
"padding": 0
},
"enabledPlugins": {
"context7@claude-plugins-official": true
},
"alwaysThinkingEnabled": true,
"effortLevel": "high",
"tui": "fullscreen",
"theme": "dark",
"verbose": true,
"remoteControlAtStartup": true,
"agentPushNotifEnabled": true,
"skipAutoPermissionPrompt": true,
"model": "sonnet"
}

37
dot_claude/skills/commit/SKILL.md Normal file
View File

@@ -0,0 +1,37 @@
---
name: commit
description: >-
Create git commits with well-formatted conventional commit messages.
disable-model-invocation: true
effort: low
model: haiku
allowed-tools:
- Glob
- Grep
- Read
- TodoWrite
- TaskCreate
- TaskUpdate
- TaskList
---
# Commit
Create commits with well-formatted conventional commit messages.
## Process
1. Check for changes (`git status`). If something is staged, consider only staged changes. Otherwise, review and stage appropriate files.
2. Determine the commit type (e.g. feat, fix, docs, refactor, ...).
3. Generate a descriptive commit message using conventional commit standards. Include scope if applicable: `type(scope): description`.
4. Add a body for complex changes explaining the *why*.
5. Execute the commit.
## Best Practices
- Keep commits atomic and focused
- Write in imperative style ("Add feature" not "Added feature")
- Explain why, not just what
- Reference issues/PRs when relevant
- Split unrelated changes into separate commits
- Do not add Claude Code as co-author to the commit

341
dot_claude/skills/sdd-code-review/SKILL.md Normal file
View File

@@ -0,0 +1,341 @@
---
name: sdd-code-review
description: >-
Review implemented code against the spec-for-change and acceptance criteria using parallel specialist subagents.
Use when asked to "review the implementation", "check the code against the spec",
"did the implementation match the spec", or "review the changes".
Also invoked internally by sdd-implement at run completion.
context: fork
model: opus
effort: high
allowed-tools:
- Agent
- AskUserQuestion
- Bash(git diff:*)
- Bash(git log:*)
- Bash(git status:*)
- Bash(git show:*)
- Glob
- Grep
- Read
- LS
---
# Spec Code Review
You are a review orchestrator. You fan out specialized reviewer subagents in parallel and
aggregate their findings into a single verdict.
**Subagent stop guard:** If you were dispatched to perform a single review task, produce
your findings and return. Do NOT invoke further subagents.
## Step 1: Gather review context
**Read the spec first.** The spec defines what the implementation must deliver — it is
the primary input for every reviewer.
1. Read the spec file completely (Goals, Functional + Non-Functional Requirements and their
Verify blocks, Design & Architectural Decisions, Edge Cases, Scope of Change, Risks & Constraints).
2. From the Design & Architectural Decisions and Scope of Change (Code) sections, identify the
source files and areas the implementation should touch. Use Read, Glob, and Grep to read the
current state of those files directly — the goal is to review what exists now, not only what changed.
3. From the **Documentation** subsection under Scope of Change, locate the architecture docs
(`docs/architecture/*`) and user docs that should have been updated, and read their current
state — the Documentation reviewer needs them.
4. Run `git status` to see which files were modified. Pass this as supplementary context.
If called by `sdd-implement`: the spec path, tasks implemented, and the list of docs updated
are passed in — use them. If called standalone: ask the user for the spec path. If no spec
exists, ask for the acceptance criteria and the files to review.
Assemble for each reviewer:
- The spec content (or acceptance criteria if no spec)
- The relevant source file contents you read
- The current contents of the docs named under Scope of Change → Documentation
- The `git status` output
## Step 2: Fan out reviewer subagents in parallel
Dispatch all applicable reviewers **at once**. Do not wait for one before dispatching
the next. Wait for all to complete before aggregating.
### Agent 1 — Developer Review
```
Task (general-purpose):
You are a senior software engineer reviewing code for quality, correctness, and spec compliance.
Do NOT modify any code. Do NOT invoke subagents.
Task: <title>
Description: <what was implemented>
Acceptance criteria:
<list>
Interface contracts:
<if applicable>
Spec / acceptance criteria:
<spec content or acceptance criteria>
Current implementation (relevant source files):
<contents of key files>
Modified files (git status):
<status output>
Review for:
- Correctness: Does the implementation satisfy every acceptance criterion?
- Spec compliance: Is every Goal (G), Functional Requirement (FR), and interface decision in
the spec satisfied? Flag any G/FR/interface that is unimplemented or only partially met.
- Scope creep: Does it touch code outside the task's scope, or add features the spec did not ask for?
- Code quality: Is code clean, readable, and maintainable? Are names descriptive? Is there unneeded complexity?
- Standards: Does it follow coding conventions existing in the project?
- Performance: Are there efficiency concerns introduced by this change?
- Scalability: Flag only if this task introduces a pattern that will actively hurt scalability at production load.
- Refactoring opportunities: Flag only if this task introduces duplication or a misplaced abstraction that will compound over time.
Produce a structured report. If a category has no findings, say so explicitly.
Output:
## Critical
<Issues that will cause failures or spec violations. Must be fixed.>
## Important
<Issues that should be fixed before merging.>
## Minor
<Small nits, style issues, optional improvements.>
## Verdict
APPROVED | FIX_AND_RECHECK | ESCALATE_TO_HUMAN
```
### Agent 2 — Quality Engineer Review
```
Task (general-purpose):
You are a QA engineer reviewing code for test coverage and correctness.
Do NOT modify any code. Do NOT invoke subagents.
Task: <title>
Acceptance criteria:
<list>
Spec / acceptance criteria:
<spec content or acceptance criteria>
Current implementation (relevant source files):
<contents of key files>
Modified files (git status):
<status output>
Review for:
- Verify-block coverage: For each Functional Requirement, is its **Verify** block actually
satisfied — is there a test or observable check matching it? Flag any FR whose Verify is unmet or untested.
- Non-functional requirements: For each NFR with a Verify bar (performance, limits,
reliability, etc.), is there evidence or a test that it is met? Flag untested NFRs.
- Test coverage: Are new/changed code paths covered? Identify uncovered branches, conditions, or functions.
- Edge cases: Are boundary conditions handled? Null/empty inputs, off-by-one, overflow, unicode, concurrency?
- Error handling: Are errors caught, propagated, and logged? Any bare catches or swallowed exceptions?
- Regression risk: Could this break existing functionality? Which existing tests are most affected?
- Missing test scenarios: List specific test cases that should be added (input → expected output where possible).
Produce a structured report. If a category has no findings, say so explicitly.
Output:
## Critical
## Important
## Minor
## Verdict
APPROVED | FIX_AND_RECHECK | ESCALATE_TO_HUMAN
```
### Agent 3 — Security Engineer Review
```
Task (general-purpose):
You are a security engineer reviewing code for vulnerabilities and data safety.
Do NOT modify any code. Do NOT invoke subagents.
If this change has no security surface (pure UI layout, config-only, documentation),
state that explicitly and output Verdict: APPROVED with no findings.
Task: <title>
Description: <what was implemented>
Spec / acceptance criteria:
<spec content or acceptance criteria>
Current implementation (relevant source files):
<contents of key files>
Modified files (git status):
<status output>
Review for:
- Vulnerabilities: Injection (SQL, command, template), XSS, CSRF, SSRF, path traversal, deserialization, insecure randomness, timing attacks.
- Data handling: Is sensitive data (PII, credentials, tokens) properly protected? Logged, cached, or serialized where it shouldn't be?
- Auth & authorization: Are auth checks present and correct? Any endpoints or operations that bypass access control?
- Input validation: Is all external input validated and sanitized? HTTP params, file uploads, env vars, config values.
- Dependencies: Are new dependencies introduced? Known CVEs? Pinned to specific versions?
- Secrets: Are there hardcoded secrets, API keys, or credentials in the diff?
Produce a structured report. Flag HIGH severity findings at the top with a clear label. If a category has no findings, say so explicitly.
Output:
## Critical
## Important
## Minor
## Verdict
APPROVED | FIX_AND_RECHECK | ESCALATE_TO_HUMAN
```
### Agent 4 — DevOps Review
```
Task (general-purpose):
You are a DevOps/platform engineer reviewing code for operational readiness.
Do NOT modify any code. Do NOT invoke subagents.
Task: <title>
Description: <what was implemented>
Spec / acceptance criteria:
<spec content or acceptance criteria>
Current implementation (relevant source files):
<contents of key files>
Modified files (git status):
<status output>
Review for:
- CI/CD impact: Will existing pipelines succeed? New build steps, dependencies, or env vars required?
- Configuration: Hardcoded values that should be env vars? Missing defaults? Config drift between environments?
- Infrastructure: New infrastructure requirements (databases, queues, storage)? Migration scripts with rollback support?
- Observability: Are metrics, logs, and traces adequate for debugging production issues? Structured log fields used?
- Rollback safety: Can this be safely reverted? Irreversible side effects (data migrations, schema changes, external API calls)?
- Resource usage: Memory, CPU, disk, or network implications? Connection pool sizing? Timeout configurations?
Produce a structured report. If a category has no findings, say so explicitly.
Output:
## Critical
## Important
## Minor
## Verdict
APPROVED | FIX_AND_RECHECK | ESCALATE_TO_HUMAN
```
### Agent 5 — UI/UX Designer Review
Dispatch only if the change touches UI code (templates, components, stylesheets, frontend
assets, accessibility attributes). Otherwise skip and note:
"UI/UX review skipped — no user-facing changes detected."
```
Task (general-purpose):
You are a UI/UX designer reviewing code for user experience quality.
Do NOT modify any code. Do NOT invoke subagents.
Task: <title>
Description: <what was implemented>
Spec / acceptance criteria:
<spec content or acceptance criteria>
Current implementation (relevant source files):
<contents of key files>
Modified files (git status):
<status output>
Review for:
- Visual consistency: Does the change follow the project's design system? Spacing, colors, and typography consistent?
- Usability: Is the interaction intuitive? Confusing states, missing feedback, or unclear labels?
- Accessibility: Proper ARIA attributes? Keyboard navigability? Sufficient color contrast? Screen reader compatibility?
- Responsive design: Works across viewport sizes? Breakpoints handled? Touch targets sized for mobile?
- Loading & error states: Loading indicators present? Error states clear and actionable? Graceful degradation?
Produce a structured report. If a category has no findings, say so explicitly.
Output:
## Critical
## Important
## Minor
## Verdict
APPROVED | FIX_AND_RECHECK | ESCALATE_TO_HUMAN
```
### Agent 6 — Documentation & Architecture Review
A peer reviewer with the **same standing as the others — no special veto**. Its findings flow
through the normal aggregation. Always dispatch it (every spec has a Documentation entry under
Scope of Change to check against).
```
Task (general-purpose):
You are a documentation reviewer checking that the architecture and user docs were updated
to match the implemented change. Do NOT modify any code or docs. Do NOT invoke subagents.
Documentation entries the spec requires (the Documentation subsection under "Scope of Change"):
<paste those entries>
Docs reported as updated (from sdd-implement, if provided):
<list, or "not provided">
Current contents of the relevant docs:
<contents of the architecture/user docs named under Scope of Change → Documentation>
What was implemented:
<summary of tasks + modified files / git status>
Review for:
- Architecture docs: Do docs/architecture/* reflect the new design — data ownership, interfaces,
patterns, and any new/changed components? Flag anything now stale or contradicted by the code.
- User docs: Are user-facing behaviours, flags, APIs, or workflows that changed documented accurately?
- Coverage: Was every documentation entry the spec required (under Scope of Change) addressed?
- Drift: Are there docs that the change made incorrect but that were left untouched?
- Accuracy: Do the updated docs actually match the code, or do they describe intended-but-absent behaviour?
Report stale, missing, or inaccurate docs as ordinary findings at the appropriate severity.
Produce a structured report. If a category has no findings, say so explicitly.
Output:
## Critical
## Important
## Minor
## Verdict
APPROVED | FIX_AND_RECHECK | ESCALATE_TO_HUMAN
```
## Step 3: Aggregate results
Once all subagents return:
1. **Overall verdict** = the worst individual verdict
(ESCALATE_TO_HUMAN > FIX_AND_RECHECK > APPROVED)
2. **Deduplicate** overlapping findings across reviewers
3. **Group by severity**: Critical → Important → Minor
## Step 4: Return results
**Called by `sdd-implement`:**
- Pass the aggregated verdict and grouped findings back into orchestration context.
- Only surface output to the user if verdict is FIX_AND_RECHECK or ESCALATE_TO_HUMAN.
- If APPROVED, stay silent and let sdd-implement continue.
**Standalone:**
- Present the full review: overall verdict, per-reviewer findings grouped by severity.
- Suggest next steps based on verdict:
- APPROVED → safe to merge / continue
- FIX_AND_RECHECK → address Critical and Important items, then re-run review
- ESCALATE_TO_HUMAN → specific issues requiring human judgment are listed above
## Principles
- All reviewer subagents run in parallel — never sequentially.
- Reviewer subagents do not invoke further subagents; they return findings and return.
- Skip a reviewer explicitly rather than silently when it doesn't apply (Security/UI may not
apply; the Documentation reviewer always applies).
- The Documentation reviewer is a peer with no special veto — its findings aggregate like any other's.
- Aggregate the worst verdict — never average or soften findings across reviewers.
- When called by sdd-implement, stay silent on APPROVED — only speak when action is needed.
- Mark findings by severity; Critical issues are never buried or omitted.

263
dot_claude/skills/sdd-implement/SKILL.md Normal file
View File

@@ -0,0 +1,263 @@
---
name: sdd-implement
description: >-
Derive an implementation plan from an approved spec-for-change and execute it task by task,
updating architecture and user docs as part of the work.
Use when asked to "implement the spec", "start implementing <feature>", "implement spec-<slug>.md",
or when the user points to a spec file to implement. Accepts a --worktree flag.
model: opus
effort: high
allowed-tools:
- Agent
- AskUserQuestion
- Bash
- EnterWorktree
- Edit
- Glob
- Grep
- Read
- Skill
- TodoWrite
- Write
- LS
---
# Spec Implement
You are an implementation orchestrator. You read an approved spec-for-change, derive your own
ordered task plan, and execute it task by task using fresh subagents — then you update the
project documentation and run review.
The spec is guardrails (*what* and *why*), not a step-by-step plan. **You are responsible for
the *how*** — think hard, break the work down yourself, and stay inside the spec's guardrails.
Never transcribe a step list from the spec; derive it.
**Stop guard:** If you were dispatched as a subagent to implement a single task, do NOT
activate this orchestration workflow. Execute the assigned task and report status only
(DONE | DONE_WITH_CONCERNS | BLOCKED | NEEDS_CONTEXT).
## Step 1: Prerequisites & arguments
1. Locate the spec file (`spec-<slug>.md` in the repo root). If none is provided and several
exist, ask which one. Read it completely (no limit/offset).
2. Parse `$ARGUMENTS` for the `--worktree` flag. Default (flag absent) is to work in the
current repo.
## Step 2: Worktree setup (only if `--worktree`)
Skip this entire step when the flag is absent.
1. **Detect existing isolation.** Compare `git rev-parse --git-dir` and
`git rev-parse --git-common-dir`; if they differ, you are already in a worktree — do NOT
nest another. Work where you are and note it.
2. **Create the worktree** on a new branch named after the slug (e.g. `sdd/<slug>`), located
as a **sibling** of the repo: `../<repo-name>-worktrees/<slug>/`. Prefer the native
`EnterWorktree` tool; fall back to `git worktree add ../<repo-name>-worktrees/<slug> -b sdd/<slug>`.
3. **Copy gitignored local config** the build needs. Discover candidates from **both
`.gitignore` and `.git/info/exclude`** (plus obvious local config in the repo root) — do not
rely on a fixed filename list. Copy the small, config-like files; examples to recognise:
`.env` / `.env.*`, service-account JSON, `.npmrc`, certs/keys, `*.tfvars`, local sqlite
databases, `.tool-versions`. Do **not** copy dependency/artifact trees (`node_modules`,
`target`, `dist`, `venv`, `.venv`, build caches) — those are reinstalled. List what you
copied, and ask the user about anything ambiguous or large rather than copying or skipping
it silently.
4. **Auto-setup:** install dependencies (detect: `package.json`→npm/pnpm/yarn, `Cargo.toml`→cargo,
`requirements.txt`/`pyproject.toml`→uv/pip/poetry, etc.).
5. **Baseline test — record, don't gate.** Run the test suite to establish a baseline.
- **Green:** note it and proceed.
- **Red:** report the failures and **ask** the user whether to proceed, investigate, or
abort. Whatever they choose, **record the set of failing tests as the baseline** so Step 6
treats only *newly* failing tests as regressions. Do not silently build on a red suite, but
do not hard-abort either.
- **Cannot run** (no test command found): note it and proceed — don't stall on it.
All subsequent steps run inside the worktree.
## Step 3: Orient, then derive the task plan
First **orient yourself in the code.** Read the files named under the spec's
**Scope of Change → Code** section yourself — a bounded skim for structure and existing
patterns, not exhaustive deep-reading. SDD specs are small by design, so holding this map in
your own context is fine; it lets you order tasks against reality and tell each executor
exactly which files to read. Only if the surface is genuinely large or unfamiliar should you
delegate this mapping to a subagent — default to reading it yourself.
Then produce an ordered task list in context (no file):
- Order by dependency: foundational work before integration before polish.
- Break work into atomic, independently-verifiable units.
- Derive acceptance criteria primarily from each Functional Requirement's **Verify block** —
those are the spec's ready-made acceptance checks. Supplement with Goals (G1…), the Verify
bars on Non-Functional Requirements (NFR1…), and Edge Cases.
- For each task, note which Goals/FRs it covers — every G and FR must appear in at least one
task — and compute its **`read_first`**: the specific files that task's executor should read
before editing (drawn from your orientation read). This is what lets executors start informed
instead of rediscovering the layout with blind `ls`/`grep` sweeps.
- Tag each task with a **model tier** for its executor: **haiku** for mechanical/boilerplate
work (renames, moves, wiring, config, repetitive edits), **sonnet** for everything else.
**Never opus for implementation tasks** — the hard reasoning is done here in the orchestrator,
so executors run on the cheaper tier.
**Record the task list with the TodoWrite tool — one todo per task — before doing anything
else.** This is a mandatory tool call and the source of truth for progress; it is *not*
satisfied by printing a markdown table. (You may also show the user a readable summary in
addition, but that does not replace the TodoWrite call.) The TodoWrite list is what keeps
progress visible during unattended runs and lets you recover your place if context is
compacted mid-execution.
Then present the plan to the user and wait for approval before executing.
## Step 4: Execute tasks in order
For each task, run 4a–4c sequentially. Keep the TodoWrite list current: exactly one task
`in_progress` at a time, mark it `completed` before moving on.
### 4a. Dispatch implementer subagent
Mark this task `in_progress` in TodoWrite, then dispatch. Give the subagent everything it needs
up front so it does not explore blindly. Fill every
field from the spec and your Step 3 orientation. Dispatch via the Agent tool with `model` set
to the task's tier from Step 3 — **`sonnet` or `haiku` only, never `opus`** for implementation
tasks:
```
Task (general-purpose):
You are implementing one task of a larger feature. Implement exactly this task — do not touch
unrelated code, do not add unrequested features.
Feature: <one-line summary of the overall change>
This is task <N> of <M>. Prior tasks produced: <what already exists / was built>.
Depends-on / enables: <what later work builds on this task>.
Task: <TASK-NN: title and description>
Covers: <G1, FR2, … — what this task fulfils>
read_first (read these before editing — the relevant existing code and patterns):
<the task's read_first file list from Step 3>
Acceptance criteria (from the FR Verify blocks + relevant NFR bars / edge cases):
<list from task plan>
Spec excerpts that bound this task:
- Requirements: <the relevant FR(s)/NFR(s)>
- Design & Architectural Decisions that apply: <paste the relevant ones>
- Guardrails — do NOT do: <the relevant Non-Goals / Constraints>
Conventions: follow the patterns already used in the read_first files.
If anything you need is missing or ambiguous, STOP and report NEEDS_CONTEXT — do not guess.
Steps:
1. Read the read_first files
2. Implement the required changes
3. Write or update tests covering the acceptance criteria
4. Run those tests, fix failures, and self-review against the acceptance criteria
Do NOT commit — the orchestrator handles commits at the end.
Report tersely — no diffs, no pasted code — with exactly these fields:
- Status: DONE | DONE_WITH_CONCERNS | BLOCKED | NEEDS_CONTEXT
- Files changed: <paths>
- Docs note: <1–2 lines on anything doc-relevant — new/renamed interfaces, behaviour, flags,
or design decisions made while implementing; "none" if nothing>
- Concerns / follow-ups: <specifics for DONE_WITH_CONCERNS/BLOCKED/NEEDS_CONTEXT; else "none">
```
### 4b. Handle status
- **DONE** → continue.
- **DONE_WITH_CONCERNS** → log concerns and continue; surface them in the final summary.
- **BLOCKED** → stop and escalate to the user with the full blocker; do not guess a workaround.
- **NEEDS_CONTEXT** → provide the missing context, re-dispatch once; if still NEEDS_CONTEXT, escalate.
### 4c. Continue
Mark the task `completed` in TodoWrite. Retain its **Docs note** and **Files changed** — Step 5
uses them to write the docs. Proceed to the next task. Do not pause between tasks.
## Step 5: Documentation & Architecture update (mandatory, done by you)
Updating docs is a core deliverable, not an afterthought.
Run it after all code tasks and before review.
Driven by the spec's **Documentation** subsection under Scope of Change plus the Docs notes you
collected, update the documentation with Edit/Write:
1. Update the architecture docs (`docs/architecture/*` or the project's equivalent) so they
match the new design.
2. Update the user-facing docs for any behaviour, flags, or APIs the user interacts with.
3. If the project has an architecture/docs index or table of contents, keep it accurate.
4. Do not invent docs structure that doesn't exist — extend what's there; if a needed doc is
genuinely absent, create it in the conventional location and note it.
Track each doc file you change so Step 6 review and the final summary can reference them. If a
required documentation entry genuinely cannot be addressed, surface it — do not silently drop it.
## Step 6: Test & review
### 6a. Run tests
Run the full test suite. If a **baseline failing set was recorded in Step 2** (worktree runs),
compare against it: only *newly* failing tests are regressions you must fix here — pre-existing
baseline failures are not yours to fix (note them, don't block on them). Fix all regressions
before reviewing. If you can't determine the test command, check for a Makefile, `package.json`
scripts, etc., or ask the user.
### 6b. Code review
Invoke the **Skill tool** with `skill: "sdd-code-review"`. This is mandatory — do NOT
self-review, do NOT substitute a general-purpose agent. Implementation size, cost, and
already-loaded context are not valid reasons to skip it.
Pass as `args`:
```
Spec file: <path to spec-<slug>.md>
Tasks implemented: <list of TASK-NN titles>
Docs updated: <list of doc files changed in Step 5>
Concerns logged: <DONE_WITH_CONCERNS items, or "none">
```
### 6c. Handle the verdict — prefer resolution
- **APPROVED** → proceed to Step 7.
- **FIX_AND_RECHECK** → you decide, and you **prefer to resolve**: dispatch a fix subagent
(Agent tool, general-purpose, `model: sonnet`) with the full findings to fix all Critical and
Important issues, then re-run `sdd-code-review` (fresh invocation). Repeat this resolve-and-recheck
loop, capped at **3 rounds**. Only **fail and report to the user** when the cap is reached
or the findings cannot be resolved — never ship Critical/Important findings silently.
- **ESCALATE_TO_HUMAN** → stop immediately and present the full findings to the user.
## Step 7: Conclude
- **If `--worktree`:** commit the work on the feature branch with a clear message, then
**print** the exact commands for the user to push and open a PR (e.g.
`git push -u origin sdd/<slug>` and `gh pr create ...`). Do **not** push or open a PR yourself.
- **Otherwise:** leave the work in the repo and suggest the user commit.
Final summary:
- Tasks completed
- Documentation files updated (Step 5)
- Code review verdict and any remaining findings (if the cap was hit)
- Concerns logged during execution (if any)
- Next steps: suggested commit / PR command (if worktree) and the spec-deletion suggestion
## Principles
- The spec is guardrails; you derive the detailed *how*. Never invent requirements not in the spec.
- Orient before planning: read the Scope-of-Change files yourself and give each executor a
`read_first` list, so subagents start informed instead of exploring blindly.
- Task order respects dependencies; foundational work comes first.
- Each implementation task runs in a fresh subagent with only the context it needs; you (the
orchestrator) stay thin by keeping their reports terse — status, files, a short docs note,
concerns — never diffs or pasted code.
- You own the codebase map and the documentation update: SDD specs are small enough that
holding that context yourself produces better docs than a cold subagent would.
- Updating architecture and user docs is mandatory (Step 5) — it is not optional and not skippable.
- Run execution cheap: implementer subagents use `sonnet` (or `haiku` for mechanical tasks),
never `opus`; the fix subagent (6c) uses `sonnet`. Only the orchestrator runs on `opus`,
because the quality-critical reasoning lives there.
- The baseline test suite is recorded, not gated: a red baseline is reported and (in worktree
mode) the user is asked; only newly failing tests count as regressions.
- Code review is a mandatory Skill call; on failure, prefer resolving and re-reviewing over failing.
- Never delete the spec — only suggest the user delete it.
- With `--worktree`, commit and print the PR command; never push or open a PR automatically.
- If blocked, stop and escalate — do not invent workarounds.

315
dot_claude/skills/sdd-plan/SKILL.md Normal file
View File

@@ -0,0 +1,315 @@
---
name: sdd-plan
description: >-
Elicit requirements, resolve unknowns via research, produce a spec-for-change, and get it approved.
Use when asked to "plan X", "spec X", "design X", "I want to build X",
"write a spec for X", or "create a spec-for-change".
model: opus
effort: high
allowed-tools:
- AskUserQuestion
- Bash(git log:*)
- Bash(git status:*)
- Bash(git rev-parse:*)
- Glob
- Grep
- Read
- WebFetch
- WebSearch
- Write
- Edit
- Skill
- TodoWrite
- LS
---
# Spec Plan
You are a technical product manager and architect. Your job is to produce a precise
**spec-for-change** that unambiguously describes *what* to build and *why* — then drive it
through agentic and human review until it is approved. The spec is the guardrails for
implementation; it describes *what* and *why*, not *how* or *in what order*.
The spec is a short-lived plan file. It lives in the repo root as `spec-<slug>.md` and is
deleted by the user (never by you) after implementation, because the durable record is the
project documentation, not the spec.
## Step 0: Seed from arguments
If `$ARGUMENTS` is non-empty, parse it before asking anything. Extract every requirement,
constraint, and decision already stated and treat them as pre-supplied answers — do NOT
re-ask for information the user already provided. Identify only the gaps that remain, then
proceed to Step 1 to fill those gaps.
## Step 1: Elicit requirements
Ask clarifying questions until you have unambiguous answers to all of:
- What problem does this solve, and why does it matter?
- What are the measurable success criteria (goals)?
- What must the system **do** (functional requirements)?
- What qualities must it meet (non-functional: performance, security, reliability,
scalability, usability, compatibility)?
- What is explicitly out of scope (non-goals / what not to do)?
- What are the known constraints (platform, dependencies, compatibility)?
- What are the important edge cases?
Do not rush to write the spec. Only proceed once you have a complete picture of the user's
intent. Technical unknowns about the codebase or external systems do not block this step —
research resolves them in Step 2.
## Step 2: Research (conditional, silent)
Invoke the **Skill tool** with `skill: "sdd-research"` — without interrupting the user — if
technical investigation is needed to write an accurate spec:
- An external API or library must be understood
- A codebase area is unfamiliar and the spec requires knowledge of existing patterns
Do NOT invoke research to resolve questions about the user's intent — those belong in Step 1.
(`sdd-research` runs in a forked context and returns only its Research Summary into yours.)
Pass as `args`:
```
Mode: internal (invoked by sdd-plan — do NOT interrupt the user; return a Research Summary and stop)
Research question: <what needs to be understood>
Decision this informs: <which part of the spec depends on this>
Cover:
- Relevant files, components, and patterns in the codebase
- External API or library behaviour (if applicable)
- Constraints and known caveats
Return a Research Summary with: Key Findings, Constraints & Risks, Open Questions, Direction.
```
**If research returns Open Questions:** surface them to the user one at a time in the same
conversational style as Step 1, and wait for answers before writing the spec.
Incorporate the research findings into the relevant spec sections.
## Step 3: Surface and resolve design decisions
A spec encodes decisions. Before drafting, find the consequential decisions the request leaves
open and let the user make them — do NOT silently bake in your own preference.
1. **Enumerate the open decisions.** From the requirements and any research, list every choice
that (a) the user did not already specify and (b) carries a real tradeoff or shapes the spec.
Look beyond the obvious architecture/approach choices for **implicit posture decisions the
user may not realize they are making** — for example whether state must survive a restart
(keep it in memory for simplicity, or persist it for durability — if the user never stated a
durability requirement, do not assume one), the error-handling philosophy (fail fast vs.
skip and continue), dependency choices, data/format, or performance targets.
2. **Ask, with tradeoffs.** Use **AskUserQuestion** to put these decisions to the user. For
each decision give 2–4 concrete options, each with its core tradeoff (the pros and cons)
stated in the option description, and put your recommended option first, marked
"(Recommended)". Batch related decisions into a single call (up to 4 questions) so you ask
once rather than repeatedly. Incorporate the answers — and, where useful, the rejected
options — into the spec's Design & Architectural Decisions and Alternatives Considered.
3. **Do not over-ask.** Only surface decisions that are genuinely consequential — those that
affect architecture, security/robustness posture, the public interface, dependencies,
data/format, or performance, or that are hard to reverse. Implementation details with no
meaningful tradeoff are the implementer's to make: leave them out of the questions and out
of the spec.
Keep the exchange design-level — no implementation steps, no file-by-file plans. Once the
decisions are made, write the spec.
## Step 4: Write the spec
Determine a short kebab-case `<slug>` for the feature. Write to `spec-<slug>.md` in the
**repo root** (run `git rev-parse --show-toplevel` to locate it; fall back to the cwd).
Use this format exactly. The spec is **guardrails — what and why, plus the decisions that
constrain implementation — never a detailed step-by-step plan.** Naming the areas expected to
change is fine; prescribing the exact sequence of edits is not — that is the implementer's job.
```markdown
# Spec: <feature name>
## Summary
<1–2 sentences: what changes and why. The elevator pitch.>
## Problem Statement
<What is broken or missing, and why it matters. The motivation for this change.>
## Current State
<Relevant parts of the system as they exist today: components, behaviour, and key files this
builds on or changes. State explicitly if this is greenfield.>
## Goals
- G1: <measurable outcome this change must achieve>
## Non-Goals
- <behaviour or feature this change deliberately does NOT address>
## Functional Requirements
<What the system must do. Every requirement is testable and carries a Verify block.>
### FR1: <requirement>
<Short description of the required behaviour.>
**Verify:** <The observable check that proves this requirement is met — a test to write, a
command to run with its expected output, or a behaviour to observe. Concrete enough to act on.>
### FR2: <requirement>
<…>
**Verify:** <…>
## Non-Functional Requirements
<Qualities and constraints: performance, security, reliability, scalability, usability,
compatibility. State a measurable bar and how it is verified, where applicable.>
### NFR1: <quality or constraint>
<Short description, with a threshold or limit where it applies.>
**Verify:** <how this is measured or confirmed>
## Design & Architectural Decisions
<The chosen approach and the decisions that constrain implementation — the shape, not the
steps. Capture ONLY the decisions that actually matter for this change. Depending on the
change these might include (non-exhaustive, illustrative): data ownership and lifecycle;
interfaces, contracts, or APIs; implementation patterns and conventions to follow; component
boundaries and responsibilities; error and failure handling strategy; concurrency model;
security/authorization model; migration or compatibility strategy. Omit what does not apply;
add what does. Use sub-headings or a decision list as suits the change.>
## Alternatives Considered
- **<Option>**: <one-sentence description>. Not chosen because: <reason>.
## Scope of Change
**Code** — indicative surface area, *not* a task list. The areas/components/files expected to
change, so the implementer has a map and the reviewer can spot scope creep:
- <component / module / path> — <why it is in scope>
**Documentation** — mandatory; kept in sync as part of implementation:
- Architecture: <which docs/architecture/* must be updated, and how>
- User docs: <which user-facing docs must be updated, and how>
## Edge Cases & Error Handling
- <case> → <expected behaviour>
## Risks & Constraints
- <hard-to-reverse decision (schema/API/migration), must-not-break, compatibility constraint, or known risk>
```
The **Documentation** subsection under Scope of Change is mandatory and non-empty —
`sdd-implement` and the docs reviewer key off it. If a change genuinely needs no doc updates,
state why explicitly there. Every Functional Requirement must carry a concrete **Verify**
block; an FR you cannot describe how to verify is not yet specified well enough.
## Step 5: Agentic review loop (owned here)
Do NOT self-review. You own this loop; `sdd-spec-review` is a pure critic that only returns
findings.
1. Invoke the **Skill tool** with `skill: "sdd-spec-review"`, passing the spec path:
```
Spec file: <path to spec-<slug>.md>
```
2. Collect the returned findings (grouped Critical / Important / Minor) and verdict.
3. **Surface the findings before you act — never fix silently.** Report to the user, for this
pass:
- the pass number and the overall verdict;
- each Critical and Important finding, one line each;
- for each, the change you will make to the spec to address it.
Format it as a short list, e.g.:
```
Review pass 1 — NEEDS_REVISION (2 Critical, 1 Important)
- [Critical] <finding> → will <fix>
- [Critical] <finding> → will <fix>
- [Important] <finding> → will <fix>
```
4. If the verdict is `NEEDS_REVISION` (any Critical or Important findings): apply the fixes you
just described to the spec for **every** Critical and Important finding, then go back to
step 1 with a **fresh** `sdd-spec-review` invocation.
5. Repeat until the reviewer returns `APPROVED` (no Critical/Important findings), capped at
**3 passes**. If issues remain after the cap, stop looping and carry them into Step 6.
**Retain each unresolved finding's full text — title, description, Impact, and Suggested
fix — verbatim from the reviewer.** You will present these richly in Step 6, so do not
reduce them to one-liners here.
## Step 6: Human approval gate (owned here)
### 6a. Present to user
Present in this order:
1. **Spec summary** — 3–5 sentences: what is being built, why, the chosen approach, and the
most important constraints/decisions.
2. **Spec path** — the file location.
3. **Resolved review findings** — a compact table of the Critical/Important findings the loop
already fixed. Keep this terse; these are done and need nothing from the user. Omit Minor
nits entirely.
| Issue | How addressed |
|-------|---------------|
| <one line> | <one line — what changed in the spec> |
4. **Open findings needing your input** — include this section **only if** findings remain
after the 3-pass cap. Do NOT put open findings in the table above, and never write
"cap reached" as their explanation — that tells the user nothing. Present each one with
enough substance to act on without re-reading the spec:
> **<finding title>**
> - What it is: <1–3 sentences describing the problem, with enough context to locate it in the spec>
> - Why it matters: <the impact — what breaks or stays ambiguous for implementation>
> - Suggested fix: <the concrete change you recommend>
(Repeat per open finding, drawn from the full finding text you retained in Step 5.) After
the list, state plainly that these are what the agentic review would still refine and that
the spec is otherwise sound — a spec need not be perfect to be implementable.
5. **Overall verdict** — APPROVED, or NEEDS_REVISION with the count of open findings.
6. **Key design decisions** — the most consequential choices made.
### 6b. Wait for response
- **User approves** → the spec is final. Stop here. Do not derive tasks. Do not implement.
Suggest the user invoke `sdd-implement` (optionally `--worktree`) to proceed.
- **User gives feedback** → you are the sole mutator of the spec: apply their changes to the
spec file, then return to **Step 6a** immediately.
- **When open findings are present**, make the choices explicit and let the user pick: (a) tell
you which open findings to address — you apply the suggested fixes (or their variant) and
re-present; (b) run another agentic review round (return to Step 5); or (c) approve as-is,
accepting the open findings. Do not re-run the agentic review automatically unless asked.
## Principles
- The spec describes *what* and *why* plus architectural decisions — never *how* or *in what
order*. The detailed implementation is the implementer's job.
- Surface consequential, tradeoff-bearing decisions to the user (with options + a
recommendation) before writing the spec — never bake in an unspecified posture, such as a
security/robustness stance, silently. Decisions with no meaningful tradeoff stay with the implementer.
- Functional and non-functional requirements both belong in the spec.
- The spec is ephemeral and lives in the repo root as `spec-<slug>.md`; never delete it.
- You own the review loop and the human gate, and you are the only skill that edits the spec.
- Always run the agentic review loop (Step 5) before the human gate — never self-review.
- On each review pass, surface the Critical/Important findings and the fix you will make for
each before applying them — the loop is visible, never silent.
- Present unresolved findings (cap reached) actionably: description, why it matters, and a
concrete suggested fix — never as a bare line whose only explanation is "cap reached".
- After human feedback, apply changes and re-present immediately (Step 6a).
- Do not implement anything, even if it seems obvious. Human approval is the gate before
implementation.

102
dot_claude/skills/sdd-research/SKILL.md Normal file
View File

@@ -0,0 +1,102 @@
---
name: sdd-research
description: >-
Resolve unknowns and gather information before making decisions.
Use when asked to "research X", "look into X", "investigate X",
"what does the codebase do around X", or "how does X API work".
Also invoked silently by sdd-plan when unknowns are detected.
context: fork
model: opus
effort: medium
allowed-tools:
- Agent
- AskUserQuestion
- Bash(git log:*)
- Bash(git grep:*)
- Glob
- Grep
- Read
- WebFetch
- WebSearch
- Write
- LS
---
# Research
You are a research specialist. Your job is to resolve unknowns and gather the information
needed to make sound decisions — either for the user directly, or silently on behalf of
a planning or agent task.
## Invocation modes
**Standalone** — the user explicitly asked you to research something.
**Internal** — invoked by another skill (e.g. `sdd-plan`) to resolve unknowns; the args
include `Mode: internal`. Do NOT interrupt the user — return the Research Summary block into
the caller's context and stop.
## Step 1: Clarify scope (standalone only)
If invoked standalone and the scope is vague (no clear decision or question driving it),
ask **one question**: "What decision will this research inform?"
Skip this step if the scope is already clear from context or arguments.
## Step 2: Research in parallel
Use all available tools concurrently. Cover all relevant dimensions:
**Codebase research** (spawn subagent):
- Find relevant files, classes, functions, and entry points
- Identify existing patterns and conventions used in this area
- Read key integration points fully (no limit/offset)
- Locate related docs/, CHANGELOG, README mentions
**External research** (if the topic involves APIs, libraries, or external systems):
- Official documentation and API references
- Version constraints and compatibility notes
- Known caveats, deprecation notices, gotchas
**Prior art** (spawn subagent):
- Similar implementations elsewhere in the codebase
- Past decisions visible in git log or comments
- Existing abstractions that could be reused
Run codebase and external research in parallel. Wait for all results before synthesizing.
## Step 3: Produce Research Summary
Synthesize all findings into a research summary:
```
## Research Summary: <topic>
### Key Findings
- <Finding with file:line or URL reference>
- <Finding with file:line or URL reference>
- [unverified] <Uncertain finding — mark explicitly>
### Constraints & Risks
- <Constraint or risk with source>
### Open Questions
- <Unresolved question that needs human input or further investigation>
### Direction
<The decision this research points toward, and why. Not implementation instructions.>
```
Only include findings you can verify directly.
Do not include assumptions, guesses, or speculation. If you have any uncertainty, mark it explicitly as `[unverified]`. Do NOT present speculation as fact.
## Step 4: Return or present
- **If called by an agent (internal):** Return the Research Summary to the invoking agent. Do NOT write a file. Do NOT speak to the user.
- **If standalone:** Present the Research Summary to the user.
## Principles
- Mark uncertain findings explicitly — never present speculation as fact.
- Run codebase and external research in parallel, not sequentially.
- Do not recommend code changes or implementation specifics — Direction means pointing toward a decision, not prescribing how to build it.
- Never produce a file unless the user explicitly asked for one.
- One clarifying question maximum; never interrogate the user.

205
dot_claude/skills/sdd-spec-review/SKILL.md Normal file
View File

@@ -0,0 +1,205 @@
---
name: sdd-spec-review
description: >-
Critique a spec-for-change with parallel specialist reviewers and return findings + a verdict.
Use when asked to "review the spec", "review this spec-for-change", "critique the spec",
or "is this spec ready to implement". Also invoked internally by sdd-plan during its review loop.
context: fork
model: opus
effort: high
allowed-tools:
- Agent
- AskUserQuestion
- Glob
- Grep
- Read
- LS
---
# Spec Review
You are a spec-review orchestrator and a **pure critic**. You fan out specialist reviewer
subagents in parallel, aggregate their findings into a single verdict, and return them.
You **never edit the spec** and you **never run a human approval gate** — those belong to
`sdd-plan`, which owns the spec and the loop. Your only output is findings + a verdict.
**Subagent stop guard:** If you were dispatched to perform a single review role, produce your
findings and return. Do NOT invoke further subagents. Do NOT edit the spec.
## Step 1: Read the spec
1. Locate the spec file. If a path was passed in, use it. If invoked standalone with no path,
look for `spec-*.md` in the repo root; if there are several, ask the user which one.
2. Read the spec completely (no limit/offset).
## Step 2: Fan out reviewer subagents in parallel
Dispatch all five reviewers **at once** (a single message, multiple Agent calls). Do not wait
for one before dispatching the next. Wait for all to return before aggregating. Pass the full
spec content to each.
Each reviewer uses the output contract below. **Every finding must be self-contained and
actionable** — enough detail that someone could fix it without re-reading the reviewer's mind.
Use this exact per-finding structure:
```
- **<short title>** — <what is wrong, missing, or ambiguous, with enough context to locate it
in the spec>. Impact: <why it matters — what breaks or stays ambiguous for the implementer>.
Suggested fix: <a concrete change to the spec that would resolve it>.
```
Group findings under three severity headers and end with a verdict:
```
## Critical
<findings that would cause the wrong thing to be built, or block implementation from starting>
## Important
<gaps an implementer could not resolve without guessing>
## Minor
<nits and optional improvements>
## Verdict
APPROVED | NEEDS_REVISION
```
**Calibrate severity — a spec is guardrails, not a formal proof or a finished implementation.**
Reserve Critical and Important for issues that genuinely block or misdirect implementation. Do
NOT escalate spec-inappropriate nits — exact stdlib error-message wording, redundant Verify
examples, or details the implementer can reasonably decide within the guardrails — to
Critical/Important; put them in Minor or omit them. If you cannot state a concrete suggested
fix for a finding, it is almost certainly not Critical or Important.
**When you dispatch each reviewer, append this entire contract verbatim to its prompt** — a
subagent only sees what you send it, so the per-finding structure and severity calibration must
be included in every reviewer prompt.
### Reviewer 1 — Completeness & Clarity
```
Task (general-purpose):
You are reviewing a spec-for-change for completeness and clarity. Do NOT edit it. Do NOT
invoke subagents. Do not praise it — find problems.
Spec content:
<full spec>
Evaluate:
- Are all sections present and non-empty (Summary, Problem Statement, Current State, Goals,
Non-Goals, Functional Requirements, Non-Functional Requirements, Design & Architectural
Decisions, Scope of Change, Edge Cases, Risks & Constraints)?
- Does every Functional Requirement carry a concrete **Verify** block?
- Is each requirement unambiguous and individually verifiable?
- Is anything important missing or left implicit?
- Is the Documentation subsection under Scope of Change concrete (names the docs to update), not hand-waved?
Output the Critical / Important / Minor / Verdict contract.
```
### Reviewer 2 — Internal Consistency
```
Task (general-purpose):
You are reviewing a spec-for-change for internal consistency. Do NOT edit it. Do NOT invoke subagents.
Spec content:
<full spec>
Evaluate:
- Do Goals, Requirements (FR/NFR), and Design & Architectural Decisions agree with each other?
- Do any Non-Goals contradict a stated Goal or Requirement?
- Do the Edge Cases and Risks & Constraints align with the Design & Architectural Decisions?
- Are there requirements with no corresponding design, or design with no driving requirement?
Output the Critical / Important / Minor / Verdict contract.
```
### Reviewer 3 — Architectural Soundness
```
Task (general-purpose):
You are a software architect reviewing a spec-for-change. Do NOT edit it. Do NOT invoke subagents.
You may read existing code with Read/Glob/Grep to check the design against reality.
Spec content:
<full spec>
Evaluate:
- Are the data-ownership, interface, and pattern decisions coherent and complete?
- Do they conflict with the Current State or established conventions in the codebase?
- Are the chosen interfaces/contracts at the right boundary? Any leaky or missing abstractions?
- Will the Design & Architectural Decisions actually satisfy the non-functional requirements?
Output the Critical / Important / Minor / Verdict contract.
```
### Reviewer 4 — Irreversibility & Risk
```
Task (general-purpose):
You are reviewing a spec-for-change for risk and irreversibility. Do NOT edit it. Do NOT invoke subagents.
Spec content:
<full spec>
Evaluate:
- Are hard-to-reverse decisions (schema changes, persisted data formats, public API/contract
changes, migrations) explicitly called out with their consequences?
- What could break existing behaviour? Are the Constraints sufficient to prevent it?
- Are there security, data-loss, or compatibility risks the spec fails to address?
- Which open questions, if any, would block implementation from starting safely?
Output the Critical / Important / Minor / Verdict contract.
```
### Reviewer 5 — Testability
```
Task (general-purpose):
You are a QA engineer reviewing a spec-for-change for testability. Do NOT edit it. Do NOT invoke subagents.
Spec content:
<full spec>
Evaluate:
- Does every Functional Requirement have a **Verify** block, and is that block concrete and
actionable (a real test, command + expected output, or observable behaviour) rather than vague?
- Are the Goals measurable?
- Are the Non-Functional Requirements expressed with verifiable bars (thresholds, limits)?
- Are the Edge Cases concrete enough to write tests from?
- Could an implementer turn the Verify blocks straight into acceptance tests, or would they have to guess?
Output the Critical / Important / Minor / Verdict contract.
```
## Step 3: Aggregate
Once all reviewers return:
1. **Overall verdict** = the worst individual verdict (`NEEDS_REVISION` > `APPROVED`).
2. **Deduplicate** overlapping findings across reviewers — when two reviewers raise the same
issue, merge them but keep the fullest description, impact, and suggested fix.
3. **Group by severity**: Critical → Important → Minor.
4. **Preserve each finding's full structure** (title, description, Impact, Suggested fix) — do
NOT shorten findings to bare titles. The caller needs the detail to act on or present them.
`APPROVED` means no Critical and no Important findings remain. Any Critical or Important
finding forces `NEEDS_REVISION`.
## Step 4: Return findings (no editing, no gate)
**Invoked by `sdd-plan` (a path was passed in):**
- Return the aggregated, deduplicated, severity-grouped findings and the overall verdict into
the caller's context. Do not address the user. Do not edit the spec.
**Standalone:**
- Present the aggregated findings and overall verdict to the user.
- Do NOT edit the spec and do NOT run an approval gate — state plainly what the user (or
`sdd-plan`) would need to fix, and stop. The user decides what to do next.
## Principles
- You are a critic, not an author: never edit the spec, never approve it on the human's behalf.
- All reviewer subagents run in parallel — never sequentially.
- Reviewer subagents do not invoke further subagents; they return findings and stop.
- Aggregate the worst verdict — never average or soften across reviewers.
- A spec with any Critical or Important finding is NEEDS_REVISION; Critical issues are never buried.

202
dot_claude/skills/skill-creator/LICENSE.txt Normal file
View File

@@ -0,0 +1,202 @@
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don't include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [yyyy] [name of copyright owner]
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

485
dot_claude/skills/skill-creator/SKILL.md Normal file
View File

@@ -0,0 +1,485 @@
---
name: skill-creator
description: Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
---
# Skill Creator
A skill for creating new skills and iteratively improving them.
At a high level, the process of creating a skill goes like this:
- Decide what you want the skill to do and roughly how it should do it
- Write a draft of the skill
- Create a few test prompts and run claude-with-access-to-the-skill on them
- Help the user evaluate the results both qualitatively and quantitatively
- While the runs happen in the background, draft some quantitative evals if there aren't any (if there are some, you can either use as is or modify if you feel something needs to change about them). Then explain them to the user (or if they already existed, explain the ones that already exist)
- Use the `eval-viewer/generate_review.py` script to show the user the results for them to look at, and also let them look at the quantitative metrics
- Rewrite the skill based on feedback from the user's evaluation of the results (and also if there are any glaring flaws that become apparent from the quantitative benchmarks)
- Repeat until you're satisfied
- Expand the test set and try again at larger scale
Your job when using this skill is to figure out where the user is in this process and then jump in and help them progress through these stages. So for instance, maybe they're like "I want to make a skill for X". You can help narrow down what they mean, write a draft, write the test cases, figure out how they want to evaluate, run all the prompts, and repeat.
On the other hand, maybe they already have a draft of the skill. In this case you can go straight to the eval/iterate part of the loop.
Of course, you should always be flexible and if the user is like "I don't need to run a bunch of evaluations, just vibe with me", you can do that instead.
Then after the skill is done (but again, the order is flexible), you can also run the skill description improver, which we have a whole separate script for, to optimize the triggering of the skill.
Cool? Cool.
## Communicating with the user
The skill creator is liable to be used by people across a wide range of familiarity with coding jargon. If you haven't heard (and how could you, it's only very recently that it started), there's a trend now where the power of Claude is inspiring plumbers to open up their terminals, parents and grandparents to google "how to install npm". On the other hand, the bulk of users are probably fairly computer-literate.
So please pay attention to context cues to understand how to phrase your communication! In the default case, just to give you some idea:
- "evaluation" and "benchmark" are borderline, but OK
- for "JSON" and "assertion" you want to see serious cues from the user that they know what those things are before using them without explaining them
It's OK to briefly explain terms if you're in doubt, and feel free to clarify terms with a short definition if you're unsure if the user will get it.
---
## Creating a skill
### Capture Intent
Start by understanding the user's intent. The current conversation might already contain a workflow the user wants to capture (e.g., they say "turn this into a skill"). If so, extract answers from the conversation history first — the tools used, the sequence of steps, corrections the user made, input/output formats observed. The user may need to fill the gaps, and should confirm before proceeding to the next step.
1. What should this skill enable Claude to do?
2. When should this skill trigger? (what user phrases/contexts)
3. What's the expected output format?
4. Should we set up test cases to verify the skill works? Skills with objectively verifiable outputs (file transforms, data extraction, code generation, fixed workflow steps) benefit from test cases. Skills with subjective outputs (writing style, art) often don't need them. Suggest the appropriate default based on the skill type, but let the user decide.
### Interview and Research
Proactively ask questions about edge cases, input/output formats, example files, success criteria, and dependencies. Wait to write test prompts until you've got this part ironed out.
Check available MCPs - if useful for research (searching docs, finding similar skills, looking up best practices), research in parallel via subagents if available, otherwise inline. Come prepared with context to reduce burden on the user.
### Write the SKILL.md
Based on the user interview, fill in these components:
- **name**: Skill identifier
- **description**: When to trigger, what it does. This is the primary triggering mechanism - include both what the skill does AND specific contexts for when to use it. All "when to use" info goes here, not in the body. Note: currently Claude has a tendency to "undertrigger" skills -- to not use them when they'd be useful. To combat this, please make the skill descriptions a little bit "pushy". So for instance, instead of "How to build a simple fast dashboard to display internal Anthropic data.", you might write "How to build a simple fast dashboard to display internal Anthropic data. Make sure to use this skill whenever the user mentions dashboards, data visualization, internal metrics, or wants to display any kind of company data, even if they don't explicitly ask for a 'dashboard.'"
- **compatibility**: Required tools, dependencies (optional, rarely needed)
- **the rest of the skill :)**
### Skill Writing Guide
#### Anatomy of a Skill
```
skill-name/
├── SKILL.md (required)
│ ├── YAML frontmatter (name, description required)
│ └── Markdown instructions
└── Bundled Resources (optional)
├── scripts/ - Executable code for deterministic/repetitive tasks
├── references/ - Docs loaded into context as needed
└── assets/ - Files used in output (templates, icons, fonts)
```
#### Progressive Disclosure
Skills use a three-level loading system:
1. **Metadata** (name + description) - Always in context (~100 words)
2. **SKILL.md body** - In context whenever skill triggers (<500 lines ideal)
3. **Bundled resources** - As needed (unlimited, scripts can execute without loading)
These word counts are approximate and you can feel free to go longer if needed.
**Key patterns:**
- Keep SKILL.md under 500 lines; if you're approaching this limit, add an additional layer of hierarchy along with clear pointers about where the model using the skill should go next to follow up.
- Reference files clearly from SKILL.md with guidance on when to read them
- For large reference files (>300 lines), include a table of contents
**Domain organization**: When a skill supports multiple domains/frameworks, organize by variant:
```
cloud-deploy/
├── SKILL.md (workflow + selection)
└── references/
├── aws.md
├── gcp.md
└── azure.md
```
Claude reads only the relevant reference file.
#### Principle of Lack of Surprise
This goes without saying, but skills must not contain malware, exploit code, or any content that could compromise system security. A skill's contents should not surprise the user in their intent if described. Don't go along with requests to create misleading skills or skills designed to facilitate unauthorized access, data exfiltration, or other malicious activities. Things like a "roleplay as an XYZ" are OK though.
#### Writing Patterns
Prefer using the imperative form in instructions.
**Defining output formats** - You can do it like this:
```markdown
## Report structure
ALWAYS use this exact template:
# [Title]
## Executive summary
## Key findings
## Recommendations
```
**Examples pattern** - It's useful to include examples. You can format them like this (but if "Input" and "Output" are in the examples you might want to deviate a little):
```markdown
## Commit message format
**Example 1:**
Input: Added user authentication with JWT tokens
Output: feat(auth): implement JWT-based authentication
```
### Writing Style
Try to explain to the model why things are important in lieu of heavy-handed musty MUSTs. Use theory of mind and try to make the skill general and not super-narrow to specific examples. Start by writing a draft and then look at it with fresh eyes and improve it.
### Test Cases
After writing the skill draft, come up with 2-3 realistic test prompts — the kind of thing a real user would actually say. Share them with the user: [you don't have to use this exact language] "Here are a few test cases I'd like to try. Do these look right, or do you want to add more?" Then run them.
Save test cases to `evals/evals.json`. Don't write assertions yet — just the prompts. You'll draft assertions in the next step while the runs are in progress.
```json
{
"skill_name": "example-skill",
"evals": [
{
"id": 1,
"prompt": "User's task prompt",
"expected_output": "Description of expected result",
"files": []
}
]
}
```
See `references/schemas.md` for the full schema (including the `assertions` field, which you'll add later).
## Running and evaluating test cases
This section is one continuous sequence — don't stop partway through. Do NOT use `/skill-test` or any other testing skill.
Put results in `<skill-name>-workspace/` as a sibling to the skill directory. Within the workspace, organize results by iteration (`iteration-1/`, `iteration-2/`, etc.) and within that, each test case gets a directory (`eval-0/`, `eval-1/`, etc.). Don't create all of this upfront — just create directories as you go.
### Step 1: Spawn all runs (with-skill AND baseline) in the same turn
For each test case, spawn two subagents in the same turn — one with the skill, one without. This is important: don't spawn the with-skill runs first and then come back for baselines later. Launch everything at once so it all finishes around the same time.
**With-skill run:**
```
Execute this task:
- Skill path: <path-to-skill>
- Task: <eval prompt>
- Input files: <eval files if any, or "none">
- Save outputs to: <workspace>/iteration-<N>/eval-<ID>/with_skill/outputs/
- Outputs to save: <what the user cares about — e.g., "the .docx file", "the final CSV">
```
**Baseline run** (same prompt, but the baseline depends on context):
- **Creating a new skill**: no skill at all. Same prompt, no skill path, save to `without_skill/outputs/`.
- **Improving an existing skill**: the old version. Before editing, snapshot the skill (`cp -r <skill-path> <workspace>/skill-snapshot/`), then point the baseline subagent at the snapshot. Save to `old_skill/outputs/`.
Write an `eval_metadata.json` for each test case (assertions can be empty for now). Give each eval a descriptive name based on what it's testing — not just "eval-0". Use this name for the directory too. If this iteration uses new or modified eval prompts, create these files for each new eval directory — don't assume they carry over from previous iterations.
```json
{
"eval_id": 0,
"eval_name": "descriptive-name-here",
"prompt": "The user's task prompt",
"assertions": []
}
```
### Step 2: While runs are in progress, draft assertions
Don't just wait for the runs to finish — you can use this time productively. Draft quantitative assertions for each test case and explain them to the user. If assertions already exist in `evals/evals.json`, review them and explain what they check.
Good assertions are objectively verifiable and have descriptive names — they should read clearly in the benchmark viewer so someone glancing at the results immediately understands what each one checks. Subjective skills (writing style, design quality) are better evaluated qualitatively — don't force assertions onto things that need human judgment.
Update the `eval_metadata.json` files and `evals/evals.json` with the assertions once drafted. Also explain to the user what they'll see in the viewer — both the qualitative outputs and the quantitative benchmark.
### Step 3: As runs complete, capture timing data
When each subagent task completes, you receive a notification containing `total_tokens` and `duration_ms`. Save this data immediately to `timing.json` in the run directory:
```json
{
"total_tokens": 84852,
"duration_ms": 23332,
"total_duration_seconds": 23.3
}
```
This is the only opportunity to capture this data — it comes through the task notification and isn't persisted elsewhere. Process each notification as it arrives rather than trying to batch them.
### Step 4: Grade, aggregate, and launch the viewer
Once all runs are done:
1. **Grade each run** — spawn a grader subagent (or grade inline) that reads `agents/grader.md` and evaluates each assertion against the outputs. Save results to `grading.json` in each run directory. The grading.json expectations array must use the fields `text`, `passed`, and `evidence` (not `name`/`met`/`details` or other variants) — the viewer depends on these exact field names. For assertions that can be checked programmatically, write and run a script rather than eyeballing it — scripts are faster, more reliable, and can be reused across iterations.
2. **Aggregate into benchmark** — run the aggregation script from the skill-creator directory:
```bash
python -m scripts.aggregate_benchmark <workspace>/iteration-N --skill-name <name>
```
This produces `benchmark.json` and `benchmark.md` with pass_rate, time, and tokens for each configuration, with mean ± stddev and the delta. If generating benchmark.json manually, see `references/schemas.md` for the exact schema the viewer expects.
Put each with_skill version before its baseline counterpart.
3. **Do an analyst pass** — read the benchmark data and surface patterns the aggregate stats might hide. See `agents/analyzer.md` (the "Analyzing Benchmark Results" section) for what to look for — things like assertions that always pass regardless of skill (non-discriminating), high-variance evals (possibly flaky), and time/token tradeoffs.
4. **Launch the viewer** with both qualitative outputs and quantitative data:
```bash
nohup python <skill-creator-path>/eval-viewer/generate_review.py \
<workspace>/iteration-N \
--skill-name "my-skill" \
--benchmark <workspace>/iteration-N/benchmark.json \
> /dev/null 2>&1 &
VIEWER_PID=$!
```
For iteration 2+, also pass `--previous-workspace <workspace>/iteration-<N-1>`.
**Cowork / headless environments:** If `webbrowser.open()` is not available or the environment has no display, use `--static <output_path>` to write a standalone HTML file instead of starting a server. Feedback will be downloaded as a `feedback.json` file when the user clicks "Submit All Reviews". After download, copy `feedback.json` into the workspace directory for the next iteration to pick up.
Note: please use generate_review.py to create the viewer; there's no need to write custom HTML.
5. **Tell the user** something like: "I've opened the results in your browser. There are two tabs — 'Outputs' lets you click through each test case and leave feedback, 'Benchmark' shows the quantitative comparison. When you're done, come back here and let me know."
### What the user sees in the viewer
The "Outputs" tab shows one test case at a time:
- **Prompt**: the task that was given
- **Output**: the files the skill produced, rendered inline where possible
- **Previous Output** (iteration 2+): collapsed section showing last iteration's output
- **Formal Grades** (if grading was run): collapsed section showing assertion pass/fail
- **Feedback**: a textbox that auto-saves as they type
- **Previous Feedback** (iteration 2+): their comments from last time, shown below the textbox
The "Benchmark" tab shows the stats summary: pass rates, timing, and token usage for each configuration, with per-eval breakdowns and analyst observations.
Navigation is via prev/next buttons or arrow keys. When done, they click "Submit All Reviews" which saves all feedback to `feedback.json`.
### Step 5: Read the feedback
When the user tells you they're done, read `feedback.json`:
```json
{
"reviews": [
{"run_id": "eval-0-with_skill", "feedback": "the chart is missing axis labels", "timestamp": "..."},
{"run_id": "eval-1-with_skill", "feedback": "", "timestamp": "..."},
{"run_id": "eval-2-with_skill", "feedback": "perfect, love this", "timestamp": "..."}
],
"status": "complete"
}
```
Empty feedback means the user thought it was fine. Focus your improvements on the test cases where the user had specific complaints.
Kill the viewer server when you're done with it:
```bash
kill $VIEWER_PID 2>/dev/null
```
---
## Improving the skill
This is the heart of the loop. You've run the test cases, the user has reviewed the results, and now you need to make the skill better based on their feedback.
### How to think about improvements
1. **Generalize from the feedback.** The big picture thing that's happening here is that we're trying to create skills that can be used a million times (maybe literally, maybe even more who knows) across many different prompts. Here you and the user are iterating on only a few examples over and over again because it helps move faster. The user knows these examples in and out and it's quick for them to assess new outputs. But if the skill you and the user are codeveloping works only for those examples, it's useless. Rather than put in fiddly overfitty changes, or oppressively constrictive MUSTs, if there's some stubborn issue, you might try branching out and using different metaphors, or recommending different patterns of working. It's relatively cheap to try and maybe you'll land on something great.
2. **Keep the prompt lean.** Remove things that aren't pulling their weight. Make sure to read the transcripts, not just the final outputs — if it looks like the skill is making the model waste a bunch of time doing things that are unproductive, you can try getting rid of the parts of the skill that are making it do that and seeing what happens.
3. **Explain the why.** Try hard to explain the **why** behind everything you're asking the model to do. Today's LLMs are *smart*. They have good theory of mind and when given a good harness can go beyond rote instructions and really make things happen. Even if the feedback from the user is terse or frustrated, try to actually understand the task and why the user is writing what they wrote, and what they actually wrote, and then transmit this understanding into the instructions. If you find yourself writing ALWAYS or NEVER in all caps, or using super rigid structures, that's a yellow flag — if possible, reframe and explain the reasoning so that the model understands why the thing you're asking for is important. That's a more humane, powerful, and effective approach.
4. **Look for repeated work across test cases.** Read the transcripts from the test runs and notice if the subagents all independently wrote similar helper scripts or took the same multi-step approach to something. If all 3 test cases resulted in the subagent writing a `create_docx.py` or a `build_chart.py`, that's a strong signal the skill should bundle that script. Write it once, put it in `scripts/`, and tell the skill to use it. This saves every future invocation from reinventing the wheel.
This task is pretty important (we are trying to create billions a year in economic value here!) and your thinking time is not the blocker; take your time and really mull things over. I'd suggest writing a draft revision and then looking at it anew and making improvements. Really do your best to get into the head of the user and understand what they want and need.
### The iteration loop
After improving the skill:
1. Apply your improvements to the skill
2. Rerun all test cases into a new `iteration-<N+1>/` directory, including baseline runs. If you're creating a new skill, the baseline is always `without_skill` (no skill) — that stays the same across iterations. If you're improving an existing skill, use your judgment on what makes sense as the baseline: the original version the user came in with, or the previous iteration.
3. Launch the reviewer with `--previous-workspace` pointing at the previous iteration
4. Wait for the user to review and tell you they're done
5. Read the new feedback, improve again, repeat
Keep going until:
- The user says they're happy
- The feedback is all empty (everything looks good)
- You're not making meaningful progress
---
## Advanced: Blind comparison
For situations where you want a more rigorous comparison between two versions of a skill (e.g., the user asks "is the new version actually better?"), there's a blind comparison system. Read `agents/comparator.md` and `agents/analyzer.md` for the details. The basic idea is: give two outputs to an independent agent without telling it which is which, and let it judge quality. Then analyze why the winner won.
This is optional, requires subagents, and most users won't need it. The human review loop is usually sufficient.
---
## Description Optimization
The description field in SKILL.md frontmatter is the primary mechanism that determines whether Claude invokes a skill. After creating or improving a skill, offer to optimize the description for better triggering accuracy.
### Step 1: Generate trigger eval queries
Create 20 eval queries — a mix of should-trigger and should-not-trigger. Save as JSON:
```json
[
{"query": "the user prompt", "should_trigger": true},
{"query": "another prompt", "should_trigger": false}
]
```
The queries must be realistic and something a Claude Code or Claude.ai user would actually type. Not abstract requests, but requests that are concrete and specific and have a good amount of detail. For instance, file paths, personal context about the user's job or situation, column names and values, company names, URLs. A little bit of backstory. Some might be in lowercase or contain abbreviations or typos or casual speech. Use a mix of different lengths, and focus on edge cases rather than making them clear-cut (the user will get a chance to sign off on them).
Bad: `"Format this data"`, `"Extract text from PDF"`, `"Create a chart"`
Good: `"ok so my boss just sent me this xlsx file (its in my downloads, called something like 'Q4 sales final FINAL v2.xlsx') and she wants me to add a column that shows the profit margin as a percentage. The revenue is in column C and costs are in column D i think"`
For the **should-trigger** queries (8-10), think about coverage. You want different phrasings of the same intent — some formal, some casual. Include cases where the user doesn't explicitly name the skill or file type but clearly needs it. Throw in some uncommon use cases and cases where this skill competes with another but should win.
For the **should-not-trigger** queries (8-10), the most valuable ones are the near-misses — queries that share keywords or concepts with the skill but actually need something different. Think adjacent domains, ambiguous phrasing where a naive keyword match would trigger but shouldn't, and cases where the query touches on something the skill does but in a context where another tool is more appropriate.
The key thing to avoid: don't make should-not-trigger queries obviously irrelevant. "Write a fibonacci function" as a negative test for a PDF skill is too easy — it doesn't test anything. The negative cases should be genuinely tricky.
### Step 2: Review with user
Present the eval set to the user for review using the HTML template:
1. Read the template from `assets/eval_review.html`
2. Replace the placeholders:
- `__EVAL_DATA_PLACEHOLDER__` → the JSON array of eval items (no quotes around it — it's a JS variable assignment)
- `__SKILL_NAME_PLACEHOLDER__` → the skill's name
- `__SKILL_DESCRIPTION_PLACEHOLDER__` → the skill's current description
3. Write to a temp file (e.g., `/tmp/eval_review_<skill-name>.html`) and open it: `open /tmp/eval_review_<skill-name>.html`
4. The user can edit queries, toggle should-trigger, add/remove entries, then click "Export Eval Set"
5. The file downloads to `~/Downloads/eval_set.json` — check the Downloads folder for the most recent version in case there are multiple (e.g., `eval_set (1).json`)
This step matters — bad eval queries lead to bad descriptions.
### Step 3: Run the optimization loop
Tell the user: "This will take some time — I'll run the optimization loop in the background and check on it periodically."
Save the eval set to the workspace, then run in the background:
```bash
python -m scripts.run_loop \
--eval-set <path-to-trigger-eval.json> \
--skill-path <path-to-skill> \
--model <model-id-powering-this-session> \
--max-iterations 5 \
--verbose
```
Use the model ID from your system prompt (the one powering the current session) so the triggering test matches what the user actually experiences.
While it runs, periodically tail the output to give the user updates on which iteration it's on and what the scores look like.
This handles the full optimization loop automatically. It splits the eval set into 60% train and 40% held-out test, evaluates the current description (running each query 3 times to get a reliable trigger rate), then calls Claude to propose improvements based on what failed. It re-evaluates each new description on both train and test, iterating up to 5 times. When it's done, it opens an HTML report in the browser showing the results per iteration and returns JSON with `best_description` — selected by test score rather than train score to avoid overfitting.
### How skill triggering works
Understanding the triggering mechanism helps design better eval queries. Skills appear in Claude's `available_skills` list with their name + description, and Claude decides whether to consult a skill based on that description. The important thing to know is that Claude only consults skills for tasks it can't easily handle on its own — simple, one-step queries like "read this PDF" may not trigger a skill even if the description matches perfectly, because Claude can handle them directly with basic tools. Complex, multi-step, or specialized queries reliably trigger skills when the description matches.
This means your eval queries should be substantive enough that Claude would actually benefit from consulting a skill. Simple queries like "read file X" are poor test cases — they won't trigger skills regardless of description quality.
### Step 4: Apply the result
Take `best_description` from the JSON output and update the skill's SKILL.md frontmatter. Show the user before/after and report the scores.
---
### Package and Present (only if `present_files` tool is available)
Check whether you have access to the `present_files` tool. If you don't, skip this step. If you do, package the skill and present the .skill file to the user:
```bash
python -m scripts.package_skill <path/to/skill-folder>
```
After packaging, direct the user to the resulting `.skill` file path so they can install it.
---
## Claude.ai-specific instructions
In Claude.ai, the core workflow is the same (draft → test → review → improve → repeat), but because Claude.ai doesn't have subagents, some mechanics change. Here's what to adapt:
**Running test cases**: No subagents means no parallel execution. For each test case, read the skill's SKILL.md, then follow its instructions to accomplish the test prompt yourself. Do them one at a time. This is less rigorous than independent subagents (you wrote the skill and you're also running it, so you have full context), but it's a useful sanity check — and the human review step compensates. Skip the baseline runs — just use the skill to complete the task as requested.
**Reviewing results**: If you can't open a browser (e.g., Claude.ai's VM has no display, or you're on a remote server), skip the browser reviewer entirely. Instead, present results directly in the conversation. For each test case, show the prompt and the output. If the output is a file the user needs to see (like a .docx or .xlsx), save it to the filesystem and tell them where it is so they can download and inspect it. Ask for feedback inline: "How does this look? Anything you'd change?"
**Benchmarking**: Skip the quantitative benchmarking — it relies on baseline comparisons which aren't meaningful without subagents. Focus on qualitative feedback from the user.
**The iteration loop**: Same as before — improve the skill, rerun the test cases, ask for feedback — just without the browser reviewer in the middle. You can still organize results into iteration directories on the filesystem if you have one.
**Description optimization**: This section requires the `claude` CLI tool (specifically `claude -p`) which is only available in Claude Code. Skip it if you're on Claude.ai.
**Blind comparison**: Requires subagents. Skip it.
**Packaging**: The `package_skill.py` script works anywhere with Python and a filesystem. On Claude.ai, you can run it and the user can download the resulting `.skill` file.
**Updating an existing skill**: The user might be asking you to update an existing skill, not create a new one. In this case:
- **Preserve the original name.** Note the skill's directory name and `name` frontmatter field -- use them unchanged. E.g., if the installed skill is `research-helper`, output `research-helper.skill` (not `research-helper-v2`).
- **Copy to a writeable location before editing.** The installed skill path may be read-only. Copy to `/tmp/skill-name/`, edit there, and package from the copy.
- **If packaging manually, stage in `/tmp/` first**, then copy to the output directory -- direct writes may fail due to permissions.
---
## Cowork-Specific Instructions
If you're in Cowork, the main things to know are:
- You have subagents, so the main workflow (spawn test cases in parallel, run baselines, grade, etc.) all works. (However, if you run into severe problems with timeouts, it's OK to run the test prompts in series rather than parallel.)
- You don't have a browser or display, so when generating the eval viewer, use `--static <output_path>` to write a standalone HTML file instead of starting a server. Then proffer a link that the user can click to open the HTML in their browser.
- For whatever reason, the Cowork setup seems to disincline Claude from generating the eval viewer after running the tests, so just to reiterate: whether you're in Cowork or in Claude Code, after running tests, you should always generate the eval viewer for the human to look at examples before revising the skill yourself and trying to make corrections, using `generate_review.py` (not writing your own boutique html code). Sorry in advance but I'm gonna go all caps here: GENERATE THE EVAL VIEWER *BEFORE* evaluating inputs yourself. You want to get them in front of the human ASAP!
- Feedback works differently: since there's no running server, the viewer's "Submit All Reviews" button will download `feedback.json` as a file. You can then read it from there (you may have to request access first).
- Packaging works — `package_skill.py` just needs Python and a filesystem.
- Description optimization (`run_loop.py` / `run_eval.py`) should work in Cowork just fine since it uses `claude -p` via subprocess, not a browser, but please save it until you've fully finished making the skill and the user agrees it's in good shape.
- **Updating an existing skill**: The user might be asking you to update an existing skill, not create a new one. Follow the update guidance in the claude.ai section above.
---
## Reference files
The agents/ directory contains instructions for specialized subagents. Read them when you need to spawn the relevant subagent.
- `agents/grader.md` — How to evaluate assertions against outputs
- `agents/comparator.md` — How to do blind A/B comparison between two outputs
- `agents/analyzer.md` — How to analyze why one version beat another
The references/ directory has additional documentation:
- `references/schemas.md` — JSON structures for evals.json, grading.json, etc.
---
Repeating one more time the core loop here for emphasis:
- Figure out what the skill is about
- Draft or edit the skill
- Run claude-with-access-to-the-skill on test prompts
- With the user, evaluate the outputs:
- Create benchmark.json and run `eval-viewer/generate_review.py` to help the user review them
- Run quantitative evals
- Repeat until you and the user are satisfied
- Package the final skill and return it to the user.
Please add steps to your TodoList, if you have such a thing, to make sure you don't forget. If you're in Cowork, please specifically put "Create evals JSON and run `eval-viewer/generate_review.py` so human can review test cases" in your TodoList to make sure it happens.
Good luck!

274
dot_claude/skills/skill-creator/agents/analyzer.md Normal file
View File

@@ -0,0 +1,274 @@
# Post-hoc Analyzer Agent
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
## Role
After the blind comparator determines a winner, the Post-hoc Analyzer "unblids" the results by examining the skills and transcripts. The goal is to extract actionable insights: what made the winner better, and how can the loser be improved?
## Inputs
You receive these parameters in your prompt:
- **winner**: "A" or "B" (from blind comparison)
- **winner_skill_path**: Path to the skill that produced the winning output
- **winner_transcript_path**: Path to the execution transcript for the winner
- **loser_skill_path**: Path to the skill that produced the losing output
- **loser_transcript_path**: Path to the execution transcript for the loser
- **comparison_result_path**: Path to the blind comparator's output JSON
- **output_path**: Where to save the analysis results
## Process
### Step 1: Read Comparison Result
1. Read the blind comparator's output at comparison_result_path
2. Note the winning side (A or B), the reasoning, and any scores
3. Understand what the comparator valued in the winning output
### Step 2: Read Both Skills
1. Read the winner skill's SKILL.md and key referenced files
2. Read the loser skill's SKILL.md and key referenced files
3. Identify structural differences:
- Instructions clarity and specificity
- Script/tool usage patterns
- Example coverage
- Edge case handling
### Step 3: Read Both Transcripts
1. Read the winner's transcript
2. Read the loser's transcript
3. Compare execution patterns:
- How closely did each follow their skill's instructions?
- What tools were used differently?
- Where did the loser diverge from optimal behavior?
- Did either encounter errors or make recovery attempts?
### Step 4: Analyze Instruction Following
For each transcript, evaluate:
- Did the agent follow the skill's explicit instructions?
- Did the agent use the skill's provided tools/scripts?
- Were there missed opportunities to leverage skill content?
- Did the agent add unnecessary steps not in the skill?
Score instruction following 1-10 and note specific issues.
### Step 5: Identify Winner Strengths
Determine what made the winner better:
- Clearer instructions that led to better behavior?
- Better scripts/tools that produced better output?
- More comprehensive examples that guided edge cases?
- Better error handling guidance?
Be specific. Quote from skills/transcripts where relevant.
### Step 6: Identify Loser Weaknesses
Determine what held the loser back:
- Ambiguous instructions that led to suboptimal choices?
- Missing tools/scripts that forced workarounds?
- Gaps in edge case coverage?
- Poor error handling that caused failures?
### Step 7: Generate Improvement Suggestions
Based on the analysis, produce actionable suggestions for improving the loser skill:
- Specific instruction changes to make
- Tools/scripts to add or modify
- Examples to include
- Edge cases to address
Prioritize by impact. Focus on changes that would have changed the outcome.
### Step 8: Write Analysis Results
Save structured analysis to `{output_path}`.
## Output Format
Write a JSON file with this structure:
```json
{
"comparison_summary": {
"winner": "A",
"winner_skill": "path/to/winner/skill",
"loser_skill": "path/to/loser/skill",
"comparator_reasoning": "Brief summary of why comparator chose winner"
},
"winner_strengths": [
"Clear step-by-step instructions for handling multi-page documents",
"Included validation script that caught formatting errors",
"Explicit guidance on fallback behavior when OCR fails"
],
"loser_weaknesses": [
"Vague instruction 'process the document appropriately' led to inconsistent behavior",
"No script for validation, agent had to improvise and made errors",
"No guidance on OCR failure, agent gave up instead of trying alternatives"
],
"instruction_following": {
"winner": {
"score": 9,
"issues": [
"Minor: skipped optional logging step"
]
},
"loser": {
"score": 6,
"issues": [
"Did not use the skill's formatting template",
"Invented own approach instead of following step 3",
"Missed the 'always validate output' instruction"
]
}
},
"improvement_suggestions": [
{
"priority": "high",
"category": "instructions",
"suggestion": "Replace 'process the document appropriately' with explicit steps: 1) Extract text, 2) Identify sections, 3) Format per template",
"expected_impact": "Would eliminate ambiguity that caused inconsistent behavior"
},
{
"priority": "high",
"category": "tools",
"suggestion": "Add validate_output.py script similar to winner skill's validation approach",
"expected_impact": "Would catch formatting errors before final output"
},
{
"priority": "medium",
"category": "error_handling",
"suggestion": "Add fallback instructions: 'If OCR fails, try: 1) different resolution, 2) image preprocessing, 3) manual extraction'",
"expected_impact": "Would prevent early failure on difficult documents"
}
],
"transcript_insights": {
"winner_execution_pattern": "Read skill -> Followed 5-step process -> Used validation script -> Fixed 2 issues -> Produced output",
"loser_execution_pattern": "Read skill -> Unclear on approach -> Tried 3 different methods -> No validation -> Output had errors"
}
}
```
## Guidelines
- **Be specific**: Quote from skills and transcripts, don't just say "instructions were unclear"
- **Be actionable**: Suggestions should be concrete changes, not vague advice
- **Focus on skill improvements**: The goal is to improve the losing skill, not critique the agent
- **Prioritize by impact**: Which changes would most likely have changed the outcome?
- **Consider causation**: Did the skill weakness actually cause the worse output, or is it incidental?
- **Stay objective**: Analyze what happened, don't editorialize
- **Think about generalization**: Would this improvement help on other evals too?
## Categories for Suggestions
Use these categories to organize improvement suggestions:
| Category | Description |
|----------|-------------|
| `instructions` | Changes to the skill's prose instructions |
| `tools` | Scripts, templates, or utilities to add/modify |
| `examples` | Example inputs/outputs to include |
| `error_handling` | Guidance for handling failures |
| `structure` | Reorganization of skill content |
| `references` | External docs or resources to add |
## Priority Levels
- **high**: Would likely change the outcome of this comparison
- **medium**: Would improve quality but may not change win/loss
- **low**: Nice to have, marginal improvement
---
# Analyzing Benchmark Results
When analyzing benchmark results, the analyzer's purpose is to **surface patterns and anomalies** across multiple runs, not suggest skill improvements.
## Role
Review all benchmark run results and generate freeform notes that help the user understand skill performance. Focus on patterns that wouldn't be visible from aggregate metrics alone.
## Inputs
You receive these parameters in your prompt:
- **benchmark_data_path**: Path to the in-progress benchmark.json with all run results
- **skill_path**: Path to the skill being benchmarked
- **output_path**: Where to save the notes (as JSON array of strings)
## Process
### Step 1: Read Benchmark Data
1. Read the benchmark.json containing all run results
2. Note the configurations tested (with_skill, without_skill)
3. Understand the run_summary aggregates already calculated
### Step 2: Analyze Per-Assertion Patterns
For each expectation across all runs:
- Does it **always pass** in both configurations? (may not differentiate skill value)
- Does it **always fail** in both configurations? (may be broken or beyond capability)
- Does it **always pass with skill but fail without**? (skill clearly adds value here)
- Does it **always fail with skill but pass without**? (skill may be hurting)
- Is it **highly variable**? (flaky expectation or non-deterministic behavior)
### Step 3: Analyze Cross-Eval Patterns
Look for patterns across evals:
- Are certain eval types consistently harder/easier?
- Do some evals show high variance while others are stable?
- Are there surprising results that contradict expectations?
### Step 4: Analyze Metrics Patterns
Look at time_seconds, tokens, tool_calls:
- Does the skill significantly increase execution time?
- Is there high variance in resource usage?
- Are there outlier runs that skew the aggregates?
### Step 5: Generate Notes
Write freeform observations as a list of strings. Each note should:
- State a specific observation
- Be grounded in the data (not speculation)
- Help the user understand something the aggregate metrics don't show
Examples:
- "Assertion 'Output is a PDF file' passes 100% in both configurations - may not differentiate skill value"
- "Eval 3 shows high variance (50% ± 40%) - run 2 had an unusual failure that may be flaky"
- "Without-skill runs consistently fail on table extraction expectations (0% pass rate)"
- "Skill adds 13s average execution time but improves pass rate by 50%"
- "Token usage is 80% higher with skill, primarily due to script output parsing"
- "All 3 without-skill runs for eval 1 produced empty output"
### Step 6: Write Notes
Save notes to `{output_path}` as a JSON array of strings:
```json
[
"Assertion 'Output is a PDF file' passes 100% in both configurations - may not differentiate skill value",
"Eval 3 shows high variance (50% ± 40%) - run 2 had an unusual failure",
"Without-skill runs consistently fail on table extraction expectations",
"Skill adds 13s average execution time but improves pass rate by 50%"
]
```
## Guidelines
**DO:**
- Report what you observe in the data
- Be specific about which evals, expectations, or runs you're referring to
- Note patterns that aggregate metrics would hide
- Provide context that helps interpret the numbers
**DO NOT:**
- Suggest improvements to the skill (that's for the improvement step, not benchmarking)
- Make subjective quality judgments ("the output was good/bad")
- Speculate about causes without evidence
- Repeat information already in the run_summary aggregates

202
dot_claude/skills/skill-creator/agents/comparator.md Normal file
View File

@@ -0,0 +1,202 @@
# Blind Comparator Agent
Compare two outputs WITHOUT knowing which skill produced them.
## Role
The Blind Comparator judges which output better accomplishes the eval task. You receive two outputs labeled A and B, but you do NOT know which skill produced which. This prevents bias toward a particular skill or approach.
Your judgment is based purely on output quality and task completion.
## Inputs
You receive these parameters in your prompt:
- **output_a_path**: Path to the first output file or directory
- **output_b_path**: Path to the second output file or directory
- **eval_prompt**: The original task/prompt that was executed
- **expectations**: List of expectations to check (optional - may be empty)
## Process
### Step 1: Read Both Outputs
1. Examine output A (file or directory)
2. Examine output B (file or directory)
3. Note the type, structure, and content of each
4. If outputs are directories, examine all relevant files inside
### Step 2: Understand the Task
1. Read the eval_prompt carefully
2. Identify what the task requires:
- What should be produced?
- What qualities matter (accuracy, completeness, format)?
- What would distinguish a good output from a poor one?
### Step 3: Generate Evaluation Rubric
Based on the task, generate a rubric with two dimensions:
**Content Rubric** (what the output contains):
| Criterion | 1 (Poor) | 3 (Acceptable) | 5 (Excellent) |
|-----------|----------|----------------|---------------|
| Correctness | Major errors | Minor errors | Fully correct |
| Completeness | Missing key elements | Mostly complete | All elements present |
| Accuracy | Significant inaccuracies | Minor inaccuracies | Accurate throughout |
**Structure Rubric** (how the output is organized):
| Criterion | 1 (Poor) | 3 (Acceptable) | 5 (Excellent) |
|-----------|----------|----------------|---------------|
| Organization | Disorganized | Reasonably organized | Clear, logical structure |
| Formatting | Inconsistent/broken | Mostly consistent | Professional, polished |
| Usability | Difficult to use | Usable with effort | Easy to use |
Adapt criteria to the specific task. For example:
- PDF form → "Field alignment", "Text readability", "Data placement"
- Document → "Section structure", "Heading hierarchy", "Paragraph flow"
- Data output → "Schema correctness", "Data types", "Completeness"
### Step 4: Evaluate Each Output Against the Rubric
For each output (A and B):
1. **Score each criterion** on the rubric (1-5 scale)
2. **Calculate dimension totals**: Content score, Structure score
3. **Calculate overall score**: Average of dimension scores, scaled to 1-10
### Step 5: Check Assertions (if provided)
If expectations are provided:
1. Check each expectation against output A
2. Check each expectation against output B
3. Count pass rates for each output
4. Use expectation scores as secondary evidence (not the primary decision factor)
### Step 6: Determine the Winner
Compare A and B based on (in priority order):
1. **Primary**: Overall rubric score (content + structure)
2. **Secondary**: Assertion pass rates (if applicable)
3. **Tiebreaker**: If truly equal, declare a TIE
Be decisive - ties should be rare. One output is usually better, even if marginally.
### Step 7: Write Comparison Results
Save results to a JSON file at the path specified (or `comparison.json` if not specified).
## Output Format
Write a JSON file with this structure:
```json
{
"winner": "A",
"reasoning": "Output A provides a complete solution with proper formatting and all required fields. Output B is missing the date field and has formatting inconsistencies.",
"rubric": {
"A": {
"content": {
"correctness": 5,
"completeness": 5,
"accuracy": 4
},
"structure": {
"organization": 4,
"formatting": 5,
"usability": 4
},
"content_score": 4.7,
"structure_score": 4.3,
"overall_score": 9.0
},
"B": {
"content": {
"correctness": 3,
"completeness": 2,
"accuracy": 3
},
"structure": {
"organization": 3,
"formatting": 2,
"usability": 3
},
"content_score": 2.7,
"structure_score": 2.7,
"overall_score": 5.4
}
},
"output_quality": {
"A": {
"score": 9,
"strengths": ["Complete solution", "Well-formatted", "All fields present"],
"weaknesses": ["Minor style inconsistency in header"]
},
"B": {
"score": 5,
"strengths": ["Readable output", "Correct basic structure"],
"weaknesses": ["Missing date field", "Formatting inconsistencies", "Partial data extraction"]
}
},
"expectation_results": {
"A": {
"passed": 4,
"total": 5,
"pass_rate": 0.80,
"details": [
{"text": "Output includes name", "passed": true},
{"text": "Output includes date", "passed": true},
{"text": "Format is PDF", "passed": true},
{"text": "Contains signature", "passed": false},
{"text": "Readable text", "passed": true}
]
},
"B": {
"passed": 3,
"total": 5,
"pass_rate": 0.60,
"details": [
{"text": "Output includes name", "passed": true},
{"text": "Output includes date", "passed": false},
{"text": "Format is PDF", "passed": true},
{"text": "Contains signature", "passed": false},
{"text": "Readable text", "passed": true}
]
}
}
}
```
If no expectations were provided, omit the `expectation_results` field entirely.
## Field Descriptions
- **winner**: "A", "B", or "TIE"
- **reasoning**: Clear explanation of why the winner was chosen (or why it's a tie)
- **rubric**: Structured rubric evaluation for each output
- **content**: Scores for content criteria (correctness, completeness, accuracy)
- **structure**: Scores for structure criteria (organization, formatting, usability)
- **content_score**: Average of content criteria (1-5)
- **structure_score**: Average of structure criteria (1-5)
- **overall_score**: Combined score scaled to 1-10
- **output_quality**: Summary quality assessment
- **score**: 1-10 rating (should match rubric overall_score)
- **strengths**: List of positive aspects
- **weaknesses**: List of issues or shortcomings
- **expectation_results**: (Only if expectations provided)
- **passed**: Number of expectations that passed
- **total**: Total number of expectations
- **pass_rate**: Fraction passed (0.0 to 1.0)
- **details**: Individual expectation results
## Guidelines
- **Stay blind**: DO NOT try to infer which skill produced which output. Judge purely on output quality.
- **Be specific**: Cite specific examples when explaining strengths and weaknesses.
- **Be decisive**: Choose a winner unless outputs are genuinely equivalent.
- **Output quality first**: Assertion scores are secondary to overall task completion.
- **Be objective**: Don't favor outputs based on style preferences; focus on correctness and completeness.
- **Explain your reasoning**: The reasoning field should make it clear why you chose the winner.
- **Handle edge cases**: If both outputs fail, pick the one that fails less badly. If both are excellent, pick the one that's marginally better.

223
dot_claude/skills/skill-creator/agents/grader.md Normal file
View File

@@ -0,0 +1,223 @@
# Grader Agent
Evaluate expectations against an execution transcript and outputs.
## Role
The Grader reviews a transcript and output files, then determines whether each expectation passes or fails. Provide clear evidence for each judgment.
You have two jobs: grade the outputs, and critique the evals themselves. A passing grade on a weak assertion is worse than useless — it creates false confidence. When you notice an assertion that's trivially satisfied, or an important outcome that no assertion checks, say so.
## Inputs
You receive these parameters in your prompt:
- **expectations**: List of expectations to evaluate (strings)
- **transcript_path**: Path to the execution transcript (markdown file)
- **outputs_dir**: Directory containing output files from execution
## Process
### Step 1: Read the Transcript
1. Read the transcript file completely
2. Note the eval prompt, execution steps, and final result
3. Identify any issues or errors documented
### Step 2: Examine Output Files
1. List files in outputs_dir
2. Read/examine each file relevant to the expectations. If outputs aren't plain text, use the inspection tools provided in your prompt — don't rely solely on what the transcript says the executor produced.
3. Note contents, structure, and quality
### Step 3: Evaluate Each Assertion
For each expectation:
1. **Search for evidence** in the transcript and outputs
2. **Determine verdict**:
- **PASS**: Clear evidence the expectation is true AND the evidence reflects genuine task completion, not just surface-level compliance
- **FAIL**: No evidence, or evidence contradicts the expectation, or the evidence is superficial (e.g., correct filename but empty/wrong content)
3. **Cite the evidence**: Quote the specific text or describe what you found
### Step 4: Extract and Verify Claims
Beyond the predefined expectations, extract implicit claims from the outputs and verify them:
1. **Extract claims** from the transcript and outputs:
- Factual statements ("The form has 12 fields")
- Process claims ("Used pypdf to fill the form")
- Quality claims ("All fields were filled correctly")
2. **Verify each claim**:
- **Factual claims**: Can be checked against the outputs or external sources
- **Process claims**: Can be verified from the transcript
- **Quality claims**: Evaluate whether the claim is justified
3. **Flag unverifiable claims**: Note claims that cannot be verified with available information
This catches issues that predefined expectations might miss.
### Step 5: Read User Notes
If `{outputs_dir}/user_notes.md` exists:
1. Read it and note any uncertainties or issues flagged by the executor
2. Include relevant concerns in the grading output
3. These may reveal problems even when expectations pass
### Step 6: Critique the Evals
After grading, consider whether the evals themselves could be improved. Only surface suggestions when there's a clear gap.
Good suggestions test meaningful outcomes — assertions that are hard to satisfy without actually doing the work correctly. Think about what makes an assertion *discriminating*: it passes when the skill genuinely succeeds and fails when it doesn't.
Suggestions worth raising:
- An assertion that passed but would also pass for a clearly wrong output (e.g., checking filename existence but not file content)
- An important outcome you observed — good or bad — that no assertion covers at all
- An assertion that can't actually be verified from the available outputs
Keep the bar high. The goal is to flag things the eval author would say "good catch" about, not to nitpick every assertion.
### Step 7: Write Grading Results
Save results to `{outputs_dir}/../grading.json` (sibling to outputs_dir).
## Grading Criteria
**PASS when**:
- The transcript or outputs clearly demonstrate the expectation is true
- Specific evidence can be cited
- The evidence reflects genuine substance, not just surface compliance (e.g., a file exists AND contains correct content, not just the right filename)
**FAIL when**:
- No evidence found for the expectation
- Evidence contradicts the expectation
- The expectation cannot be verified from available information
- The evidence is superficial — the assertion is technically satisfied but the underlying task outcome is wrong or incomplete
- The output appears to meet the assertion by coincidence rather than by actually doing the work
**When uncertain**: The burden of proof to pass is on the expectation.
### Step 8: Read Executor Metrics and Timing
1. If `{outputs_dir}/metrics.json` exists, read it and include in grading output
2. If `{outputs_dir}/../timing.json` exists, read it and include timing data
## Output Format
Write a JSON file with this structure:
```json
{
"expectations": [
{
"text": "The output includes the name 'John Smith'",
"passed": true,
"evidence": "Found in transcript Step 3: 'Extracted names: John Smith, Sarah Johnson'"
},
{
"text": "The spreadsheet has a SUM formula in cell B10",
"passed": false,
"evidence": "No spreadsheet was created. The output was a text file."
},
{
"text": "The assistant used the skill's OCR script",
"passed": true,
"evidence": "Transcript Step 2 shows: 'Tool: Bash - python ocr_script.py image.png'"
}
],
"summary": {
"passed": 2,
"failed": 1,
"total": 3,
"pass_rate": 0.67
},
"execution_metrics": {
"tool_calls": {
"Read": 5,
"Write": 2,
"Bash": 8
},
"total_tool_calls": 15,
"total_steps": 6,
"errors_encountered": 0,
"output_chars": 12450,
"transcript_chars": 3200
},
"timing": {
"executor_duration_seconds": 165.0,
"grader_duration_seconds": 26.0,
"total_duration_seconds": 191.0
},
"claims": [
{
"claim": "The form has 12 fillable fields",
"type": "factual",
"verified": true,
"evidence": "Counted 12 fields in field_info.json"
},
{
"claim": "All required fields were populated",
"type": "quality",
"verified": false,
"evidence": "Reference section was left blank despite data being available"
}
],
"user_notes_summary": {
"uncertainties": ["Used 2023 data, may be stale"],
"needs_review": [],
"workarounds": ["Fell back to text overlay for non-fillable fields"]
},
"eval_feedback": {
"suggestions": [
{
"assertion": "The output includes the name 'John Smith'",
"reason": "A hallucinated document that mentions the name would also pass — consider checking it appears as the primary contact with matching phone and email from the input"
},
{
"reason": "No assertion checks whether the extracted phone numbers match the input — I observed incorrect numbers in the output that went uncaught"
}
],
"overall": "Assertions check presence but not correctness. Consider adding content verification."
}
}
```
## Field Descriptions
- **expectations**: Array of graded expectations
- **text**: The original expectation text
- **passed**: Boolean - true if expectation passes
- **evidence**: Specific quote or description supporting the verdict
- **summary**: Aggregate statistics
- **passed**: Count of passed expectations
- **failed**: Count of failed expectations
- **total**: Total expectations evaluated
- **pass_rate**: Fraction passed (0.0 to 1.0)
- **execution_metrics**: Copied from executor's metrics.json (if available)
- **output_chars**: Total character count of output files (proxy for tokens)
- **transcript_chars**: Character count of transcript
- **timing**: Wall clock timing from timing.json (if available)
- **executor_duration_seconds**: Time spent in executor subagent
- **total_duration_seconds**: Total elapsed time for the run
- **claims**: Extracted and verified claims from the output
- **claim**: The statement being verified
- **type**: "factual", "process", or "quality"
- **verified**: Boolean - whether the claim holds
- **evidence**: Supporting or contradicting evidence
- **user_notes_summary**: Issues flagged by the executor
- **uncertainties**: Things the executor wasn't sure about
- **needs_review**: Items requiring human attention
- **workarounds**: Places where the skill didn't work as expected
- **eval_feedback**: Improvement suggestions for the evals (only when warranted)
- **suggestions**: List of concrete suggestions, each with a `reason` and optionally an `assertion` it relates to
- **overall**: Brief assessment — can be "No suggestions, evals look solid" if nothing to flag
## Guidelines
- **Be objective**: Base verdicts on evidence, not assumptions
- **Be specific**: Quote the exact text that supports your verdict
- **Be thorough**: Check both transcript and output files
- **Be consistent**: Apply the same standard to each expectation
- **Explain failures**: Make it clear why evidence was insufficient
- **No partial credit**: Each expectation is pass or fail, not partial

146
dot_claude/skills/skill-creator/assets/eval_review.html Normal file
View File

@@ -0,0 +1,146 @@
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Eval Set Review - __SKILL_NAME_PLACEHOLDER__</title>
<link rel="preconnect" href="https://fonts.googleapis.com">
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
<link href="https://fonts.googleapis.com/css2?family=Poppins:wght@500;600&family=Lora:wght@400;500&display=swap" rel="stylesheet">
<style>
* { box-sizing: border-box; margin: 0; padding: 0; }
body { font-family: 'Lora', Georgia, serif; background: #faf9f5; padding: 2rem; color: #141413; }
h1 { font-family: 'Poppins', sans-serif; margin-bottom: 0.5rem; font-size: 1.5rem; }
.description { color: #b0aea5; margin-bottom: 1.5rem; font-style: italic; max-width: 900px; }
.controls { margin-bottom: 1rem; display: flex; gap: 0.5rem; }
.btn { font-family: 'Poppins', sans-serif; padding: 0.5rem 1rem; border: none; border-radius: 6px; cursor: pointer; font-size: 0.875rem; font-weight: 500; }
.btn-add { background: #6a9bcc; color: white; }
.btn-add:hover { background: #5889b8; }
.btn-export { background: #d97757; color: white; }
.btn-export:hover { background: #c4613f; }
table { width: 100%; max-width: 1100px; border-collapse: collapse; background: white; border-radius: 6px; overflow: hidden; box-shadow: 0 1px 3px rgba(0,0,0,0.08); }
th { font-family: 'Poppins', sans-serif; background: #141413; color: #faf9f5; padding: 0.75rem 1rem; text-align: left; font-size: 0.875rem; }
td { padding: 0.75rem 1rem; border-bottom: 1px solid #e8e6dc; vertical-align: top; }
tr:nth-child(even) td { background: #faf9f5; }
tr:hover td { background: #f3f1ea; }
.section-header td { background: #e8e6dc; font-family: 'Poppins', sans-serif; font-weight: 500; font-size: 0.8rem; color: #141413; text-transform: uppercase; letter-spacing: 0.05em; }
.query-input { width: 100%; padding: 0.4rem; border: 1px solid #e8e6dc; border-radius: 4px; font-size: 0.875rem; font-family: 'Lora', Georgia, serif; resize: vertical; min-height: 60px; }
.query-input:focus { outline: none; border-color: #d97757; box-shadow: 0 0 0 2px rgba(217,119,87,0.15); }
.toggle { position: relative; display: inline-block; width: 44px; height: 24px; }
.toggle input { opacity: 0; width: 0; height: 0; }
.toggle .slider { position: absolute; inset: 0; background: #b0aea5; border-radius: 24px; cursor: pointer; transition: 0.2s; }
.toggle .slider::before { content: ""; position: absolute; width: 18px; height: 18px; left: 3px; bottom: 3px; background: white; border-radius: 50%; transition: 0.2s; }
.toggle input:checked + .slider { background: #d97757; }
.toggle input:checked + .slider::before { transform: translateX(20px); }
.btn-delete { background: #c44; color: white; padding: 0.3rem 0.6rem; border: none; border-radius: 4px; cursor: pointer; font-size: 0.75rem; font-family: 'Poppins', sans-serif; }
.btn-delete:hover { background: #a33; }
.summary { margin-top: 1rem; color: #b0aea5; font-size: 0.875rem; }
</style>
</head>
<body>
<h1>Eval Set Review: <span id="skill-name">__SKILL_NAME_PLACEHOLDER__</span></h1>
<p class="description">Current description: <span id="skill-desc">__SKILL_DESCRIPTION_PLACEHOLDER__</span></p>
<div class="controls">
<button class="btn btn-add" onclick="addRow()">+ Add Query</button>
<button class="btn btn-export" onclick="exportEvalSet()">Export Eval Set</button>
</div>
<table>
<thead>
<tr>
<th style="width:65%">Query</th>
<th style="width:18%">Should Trigger</th>
<th style="width:10%">Actions</th>
</tr>
</thead>
<tbody id="eval-body"></tbody>
</table>
<p class="summary" id="summary"></p>
<script>
const EVAL_DATA = __EVAL_DATA_PLACEHOLDER__;
let evalItems = [...EVAL_DATA];
function render() {
const tbody = document.getElementById('eval-body');
tbody.innerHTML = '';
// Sort: should-trigger first, then should-not-trigger
const sorted = evalItems
.map((item, origIdx) => ({ ...item, origIdx }))
.sort((a, b) => (b.should_trigger ? 1 : 0) - (a.should_trigger ? 1 : 0));
let lastGroup = null;
sorted.forEach(item => {
const group = item.should_trigger ? 'trigger' : 'no-trigger';
if (group !== lastGroup) {
const headerRow = document.createElement('tr');
headerRow.className = 'section-header';
headerRow.innerHTML = `<td colspan="3">${item.should_trigger ? 'Should Trigger' : 'Should NOT Trigger'}</td>`;
tbody.appendChild(headerRow);
lastGroup = group;
}
const idx = item.origIdx;
const tr = document.createElement('tr');
tr.innerHTML = `
<td><textarea class="query-input" onchange="updateQuery(${idx}, this.value)">${escapeHtml(item.query)}</textarea></td>
<td>
<label class="toggle">
<input type="checkbox" ${item.should_trigger ? 'checked' : ''} onchange="updateTrigger(${idx}, this.checked)">
<span class="slider"></span>
</label>
<span style="margin-left:8px;font-size:0.8rem;color:#b0aea5">${item.should_trigger ? 'Yes' : 'No'}</span>
</td>
<td><button class="btn-delete" onclick="deleteRow(${idx})">Delete</button></td>
`;
tbody.appendChild(tr);
});
updateSummary();
}
function escapeHtml(text) {
const div = document.createElement('div');
div.textContent = text;
return div.innerHTML;
}
function updateQuery(idx, value) { evalItems[idx].query = value; updateSummary(); }
function updateTrigger(idx, value) { evalItems[idx].should_trigger = value; render(); }
function deleteRow(idx) { evalItems.splice(idx, 1); render(); }
function addRow() {
evalItems.push({ query: '', should_trigger: true });
render();
const inputs = document.querySelectorAll('.query-input');
inputs[inputs.length - 1].focus();
}
function updateSummary() {
const trigger = evalItems.filter(i => i.should_trigger).length;
const noTrigger = evalItems.filter(i => !i.should_trigger).length;
document.getElementById('summary').textContent =
`${evalItems.length} queries total: ${trigger} should trigger, ${noTrigger} should not trigger`;
}
function exportEvalSet() {
const valid = evalItems.filter(i => i.query.trim() !== '');
const data = valid.map(i => ({ query: i.query.trim(), should_trigger: i.should_trigger }));
const blob = new Blob([JSON.stringify(data, null, 2)], { type: 'application/json' });
const url = URL.createObjectURL(blob);
const a = document.createElement('a');
a.href = url;
a.download = 'eval_set.json';
document.body.appendChild(a);
a.click();
document.body.removeChild(a);
URL.revokeObjectURL(url);
}
render();
</script>
</body>
</html>

471
dot_claude/skills/skill-creator/eval-viewer/generate_review.py Normal file
View File

@@ -0,0 +1,471 @@
#!/usr/bin/env python3
"""Generate and serve a review page for eval results.
Reads the workspace directory, discovers runs (directories with outputs/),
embeds all output data into a self-contained HTML page, and serves it via
a tiny HTTP server. Feedback auto-saves to feedback.json in the workspace.
Usage:
python generate_review.py <workspace-path> [--port PORT] [--skill-name NAME]
python generate_review.py <workspace-path> --previous-feedback /path/to/old/feedback.json
No dependencies beyond the Python stdlib are required.
"""
import argparse
import base64
import json
import mimetypes
import os
import re
import signal
import subprocess
import sys
import time
import webbrowser
from functools import partial
from http.server import HTTPServer, BaseHTTPRequestHandler
from pathlib import Path
# Files to exclude from output listings
METADATA_FILES = {"transcript.md", "user_notes.md", "metrics.json"}
# Extensions we render as inline text
TEXT_EXTENSIONS = {
".txt", ".md", ".json", ".csv", ".py", ".js", ".ts", ".tsx", ".jsx",
".yaml", ".yml", ".xml", ".html", ".css", ".sh", ".rb", ".go", ".rs",
".java", ".c", ".cpp", ".h", ".hpp", ".sql", ".r", ".toml",
}
# Extensions we render as inline images
IMAGE_EXTENSIONS = {".png", ".jpg", ".jpeg", ".gif", ".svg", ".webp"}
# MIME type overrides for common types
MIME_OVERRIDES = {
".svg": "image/svg+xml",
".xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
".docx": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
".pptx": "application/vnd.openxmlformats-officedocument.presentationml.presentation",
}
def get_mime_type(path: Path) -> str:
ext = path.suffix.lower()
if ext in MIME_OVERRIDES:
return MIME_OVERRIDES[ext]
mime, _ = mimetypes.guess_type(str(path))
return mime or "application/octet-stream"
def find_runs(workspace: Path) -> list[dict]:
"""Recursively find directories that contain an outputs/ subdirectory."""
runs: list[dict] = []
_find_runs_recursive(workspace, workspace, runs)
runs.sort(key=lambda r: (r.get("eval_id", float("inf")), r["id"]))
return runs
def _find_runs_recursive(root: Path, current: Path, runs: list[dict]) -> None:
if not current.is_dir():
return
outputs_dir = current / "outputs"
if outputs_dir.is_dir():
run = build_run(root, current)
if run:
runs.append(run)
return
skip = {"node_modules", ".git", "__pycache__", "skill", "inputs"}
for child in sorted(current.iterdir()):
if child.is_dir() and child.name not in skip:
_find_runs_recursive(root, child, runs)
def build_run(root: Path, run_dir: Path) -> dict | None:
"""Build a run dict with prompt, outputs, and grading data."""
prompt = ""
eval_id = None
# Try eval_metadata.json
for candidate in [run_dir / "eval_metadata.json", run_dir.parent / "eval_metadata.json"]:
if candidate.exists():
try:
metadata = json.loads(candidate.read_text())
prompt = metadata.get("prompt", "")
eval_id = metadata.get("eval_id")
except (json.JSONDecodeError, OSError):
pass
if prompt:
break
# Fall back to transcript.md
if not prompt:
for candidate in [run_dir / "transcript.md", run_dir / "outputs" / "transcript.md"]:
if candidate.exists():
try:
text = candidate.read_text()
match = re.search(r"## Eval Prompt\n\n([\s\S]*?)(?=\n##|$)", text)
if match:
prompt = match.group(1).strip()
except OSError:
pass
if prompt:
break
if not prompt:
prompt = "(No prompt found)"
run_id = str(run_dir.relative_to(root)).replace("/", "-").replace("\\", "-")
# Collect output files
outputs_dir = run_dir / "outputs"
output_files: list[dict] = []
if outputs_dir.is_dir():
for f in sorted(outputs_dir.iterdir()):
if f.is_file() and f.name not in METADATA_FILES:
output_files.append(embed_file(f))
# Load grading if present
grading = None
for candidate in [run_dir / "grading.json", run_dir.parent / "grading.json"]:
if candidate.exists():
try:
grading = json.loads(candidate.read_text())
except (json.JSONDecodeError, OSError):
pass
if grading:
break
return {
"id": run_id,
"prompt": prompt,
"eval_id": eval_id,
"outputs": output_files,
"grading": grading,
}
def embed_file(path: Path) -> dict:
"""Read a file and return an embedded representation."""
ext = path.suffix.lower()
mime = get_mime_type(path)
if ext in TEXT_EXTENSIONS:
try:
content = path.read_text(errors="replace")
except OSError:
content = "(Error reading file)"
return {
"name": path.name,
"type": "text",
"content": content,
}
elif ext in IMAGE_EXTENSIONS:
try:
raw = path.read_bytes()
b64 = base64.b64encode(raw).decode("ascii")
except OSError:
return {"name": path.name, "type": "error", "content": "(Error reading file)"}
return {
"name": path.name,
"type": "image",
"mime": mime,
"data_uri": f"data:{mime};base64,{b64}",
}
elif ext == ".pdf":
try:
raw = path.read_bytes()
b64 = base64.b64encode(raw).decode("ascii")
except OSError:
return {"name": path.name, "type": "error", "content": "(Error reading file)"}
return {
"name": path.name,
"type": "pdf",
"data_uri": f"data:{mime};base64,{b64}",
}
elif ext == ".xlsx":
try:
raw = path.read_bytes()
b64 = base64.b64encode(raw).decode("ascii")
except OSError:
return {"name": path.name, "type": "error", "content": "(Error reading file)"}
return {
"name": path.name,
"type": "xlsx",
"data_b64": b64,
}
else:
# Binary / unknown — base64 download link
try:
raw = path.read_bytes()
b64 = base64.b64encode(raw).decode("ascii")
except OSError:
return {"name": path.name, "type": "error", "content": "(Error reading file)"}
return {
"name": path.name,
"type": "binary",
"mime": mime,
"data_uri": f"data:{mime};base64,{b64}",
}
def load_previous_iteration(workspace: Path) -> dict[str, dict]:
"""Load previous iteration's feedback and outputs.
Returns a map of run_id -> {"feedback": str, "outputs": list[dict]}.
"""
result: dict[str, dict] = {}
# Load feedback
feedback_map: dict[str, str] = {}
feedback_path = workspace / "feedback.json"
if feedback_path.exists():
try:
data = json.loads(feedback_path.read_text())
feedback_map = {
r["run_id"]: r["feedback"]
for r in data.get("reviews", [])
if r.get("feedback", "").strip()
}
except (json.JSONDecodeError, OSError, KeyError):
pass
# Load runs (to get outputs)
prev_runs = find_runs(workspace)
for run in prev_runs:
result[run["id"]] = {
"feedback": feedback_map.get(run["id"], ""),
"outputs": run.get("outputs", []),
}
# Also add feedback for run_ids that had feedback but no matching run
for run_id, fb in feedback_map.items():
if run_id not in result:
result[run_id] = {"feedback": fb, "outputs": []}
return result
def generate_html(
runs: list[dict],
skill_name: str,
previous: dict[str, dict] | None = None,
benchmark: dict | None = None,
) -> str:
"""Generate the complete standalone HTML page with embedded data."""
template_path = Path(__file__).parent / "viewer.html"
template = template_path.read_text()
# Build previous_feedback and previous_outputs maps for the template
previous_feedback: dict[str, str] = {}
previous_outputs: dict[str, list[dict]] = {}
if previous:
for run_id, data in previous.items():
if data.get("feedback"):
previous_feedback[run_id] = data["feedback"]
if data.get("outputs"):
previous_outputs[run_id] = data["outputs"]
embedded = {
"skill_name": skill_name,
"runs": runs,
"previous_feedback": previous_feedback,
"previous_outputs": previous_outputs,
}
if benchmark:
embedded["benchmark"] = benchmark
data_json = json.dumps(embedded)
return template.replace("/*__EMBEDDED_DATA__*/", f"const EMBEDDED_DATA = {data_json};")
# ---------------------------------------------------------------------------
# HTTP server (stdlib only, zero dependencies)
# ---------------------------------------------------------------------------
def _kill_port(port: int) -> None:
"""Kill any process listening on the given port."""
try:
result = subprocess.run(
["lsof", "-ti", f":{port}"],
capture_output=True, text=True, timeout=5,
)
for pid_str in result.stdout.strip().split("\n"):
if pid_str.strip():
try:
os.kill(int(pid_str.strip()), signal.SIGTERM)
except (ProcessLookupError, ValueError):
pass
if result.stdout.strip():
time.sleep(0.5)
except subprocess.TimeoutExpired:
pass
except FileNotFoundError:
print("Note: lsof not found, cannot check if port is in use", file=sys.stderr)
class ReviewHandler(BaseHTTPRequestHandler):
"""Serves the review HTML and handles feedback saves.
Regenerates the HTML on each page load so that refreshing the browser
picks up new eval outputs without restarting the server.
"""
def __init__(
self,
workspace: Path,
skill_name: str,
feedback_path: Path,
previous: dict[str, dict],
benchmark_path: Path | None,
*args,
**kwargs,
):
self.workspace = workspace
self.skill_name = skill_name
self.feedback_path = feedback_path
self.previous = previous
self.benchmark_path = benchmark_path
super().__init__(*args, **kwargs)
def do_GET(self) -> None:
if self.path == "/" or self.path == "/index.html":
# Regenerate HTML on each request (re-scans workspace for new outputs)
runs = find_runs(self.workspace)
benchmark = None
if self.benchmark_path and self.benchmark_path.exists():
try:
benchmark = json.loads(self.benchmark_path.read_text())
except (json.JSONDecodeError, OSError):
pass
html = generate_html(runs, self.skill_name, self.previous, benchmark)
content = html.encode("utf-8")
self.send_response(200)
self.send_header("Content-Type", "text/html; charset=utf-8")
self.send_header("Content-Length", str(len(content)))
self.end_headers()
self.wfile.write(content)
elif self.path == "/api/feedback":
data = b"{}"
if self.feedback_path.exists():
data = self.feedback_path.read_bytes()
self.send_response(200)
self.send_header("Content-Type", "application/json")
self.send_header("Content-Length", str(len(data)))
self.end_headers()
self.wfile.write(data)
else:
self.send_error(404)
def do_POST(self) -> None:
if self.path == "/api/feedback":
length = int(self.headers.get("Content-Length", 0))
body = self.rfile.read(length)
try:
data = json.loads(body)
if not isinstance(data, dict) or "reviews" not in data:
raise ValueError("Expected JSON object with 'reviews' key")
self.feedback_path.write_text(json.dumps(data, indent=2) + "\n")
resp = b'{"ok":true}'
self.send_response(200)
except (json.JSONDecodeError, OSError, ValueError) as e:
resp = json.dumps({"error": str(e)}).encode()
self.send_response(500)
self.send_header("Content-Type", "application/json")
self.send_header("Content-Length", str(len(resp)))
self.end_headers()
self.wfile.write(resp)
else:
self.send_error(404)
def log_message(self, format: str, *args: object) -> None:
# Suppress request logging to keep terminal clean
pass
def main() -> None:
parser = argparse.ArgumentParser(description="Generate and serve eval review")
parser.add_argument("workspace", type=Path, help="Path to workspace directory")
parser.add_argument("--port", "-p", type=int, default=3117, help="Server port (default: 3117)")
parser.add_argument("--skill-name", "-n", type=str, default=None, help="Skill name for header")
parser.add_argument(
"--previous-workspace", type=Path, default=None,
help="Path to previous iteration's workspace (shows old outputs and feedback as context)",
)
parser.add_argument(
"--benchmark", type=Path, default=None,
help="Path to benchmark.json to show in the Benchmark tab",
)
parser.add_argument(
"--static", "-s", type=Path, default=None,
help="Write standalone HTML to this path instead of starting a server",
)
args = parser.parse_args()
workspace = args.workspace.resolve()
if not workspace.is_dir():
print(f"Error: {workspace} is not a directory", file=sys.stderr)
sys.exit(1)
runs = find_runs(workspace)
if not runs:
print(f"No runs found in {workspace}", file=sys.stderr)
sys.exit(1)
skill_name = args.skill_name or workspace.name.replace("-workspace", "")
feedback_path = workspace / "feedback.json"
previous: dict[str, dict] = {}
if args.previous_workspace:
previous = load_previous_iteration(args.previous_workspace.resolve())
benchmark_path = args.benchmark.resolve() if args.benchmark else None
benchmark = None
if benchmark_path and benchmark_path.exists():
try:
benchmark = json.loads(benchmark_path.read_text())
except (json.JSONDecodeError, OSError):
pass
if args.static:
html = generate_html(runs, skill_name, previous, benchmark)
args.static.parent.mkdir(parents=True, exist_ok=True)
args.static.write_text(html)
print(f"\n Static viewer written to: {args.static}\n")
sys.exit(0)
# Kill any existing process on the target port
port = args.port
_kill_port(port)
handler = partial(ReviewHandler, workspace, skill_name, feedback_path, previous, benchmark_path)
try:
server = HTTPServer(("127.0.0.1", port), handler)
except OSError:
# Port still in use after kill attempt — find a free one
server = HTTPServer(("127.0.0.1", 0), handler)
port = server.server_address[1]
url = f"http://localhost:{port}"
print(f"\n Eval Viewer")
print(f" ─────────────────────────────────")
print(f" URL: {url}")
print(f" Workspace: {workspace}")
print(f" Feedback: {feedback_path}")
if previous:
print(f" Previous: {args.previous_workspace} ({len(previous)} runs)")
if benchmark_path:
print(f" Benchmark: {benchmark_path}")
print(f"\n Press Ctrl+C to stop.\n")
webbrowser.open(url)
try:
server.serve_forever()
except KeyboardInterrupt:
print("\nStopped.")
server.server_close()
if __name__ == "__main__":
main()

1325
dot_claude/skills/skill-creator/eval-viewer/viewer.html Normal file
View File

File diff suppressed because it is too large Load Diff

430
dot_claude/skills/skill-creator/references/schemas.md Normal file
View File

@@ -0,0 +1,430 @@
# JSON Schemas
This document defines the JSON schemas used by skill-creator.
---
## evals.json
Defines the evals for a skill. Located at `evals/evals.json` within the skill directory.
```json
{
"skill_name": "example-skill",
"evals": [
{
"id": 1,
"prompt": "User's example prompt",
"expected_output": "Description of expected result",
"files": ["evals/files/sample1.pdf"],
"expectations": [
"The output includes X",
"The skill used script Y"
]
}
]
}
```
**Fields:**
- `skill_name`: Name matching the skill's frontmatter
- `evals[].id`: Unique integer identifier
- `evals[].prompt`: The task to execute
- `evals[].expected_output`: Human-readable description of success
- `evals[].files`: Optional list of input file paths (relative to skill root)
- `evals[].expectations`: List of verifiable statements
---
## history.json
Tracks version progression in Improve mode. Located at workspace root.
```json
{
"started_at": "2026-01-15T10:30:00Z",
"skill_name": "pdf",
"current_best": "v2",
"iterations": [
{
"version": "v0",
"parent": null,
"expectation_pass_rate": 0.65,
"grading_result": "baseline",
"is_current_best": false
},
{
"version": "v1",
"parent": "v0",
"expectation_pass_rate": 0.75,
"grading_result": "won",
"is_current_best": false
},
{
"version": "v2",
"parent": "v1",
"expectation_pass_rate": 0.85,
"grading_result": "won",
"is_current_best": true
}
]
}
```
**Fields:**
- `started_at`: ISO timestamp of when improvement started
- `skill_name`: Name of the skill being improved
- `current_best`: Version identifier of the best performer
- `iterations[].version`: Version identifier (v0, v1, ...)
- `iterations[].parent`: Parent version this was derived from
- `iterations[].expectation_pass_rate`: Pass rate from grading
- `iterations[].grading_result`: "baseline", "won", "lost", or "tie"
- `iterations[].is_current_best`: Whether this is the current best version
---
## grading.json
Output from the grader agent. Located at `<run-dir>/grading.json`.
```json
{
"expectations": [
{
"text": "The output includes the name 'John Smith'",
"passed": true,
"evidence": "Found in transcript Step 3: 'Extracted names: John Smith, Sarah Johnson'"
},
{
"text": "The spreadsheet has a SUM formula in cell B10",
"passed": false,
"evidence": "No spreadsheet was created. The output was a text file."
}
],
"summary": {
"passed": 2,
"failed": 1,
"total": 3,
"pass_rate": 0.67
},
"execution_metrics": {
"tool_calls": {
"Read": 5,
"Write": 2,
"Bash": 8
},
"total_tool_calls": 15,
"total_steps": 6,
"errors_encountered": 0,
"output_chars": 12450,
"transcript_chars": 3200
},
"timing": {
"executor_duration_seconds": 165.0,
"grader_duration_seconds": 26.0,
"total_duration_seconds": 191.0
},
"claims": [
{
"claim": "The form has 12 fillable fields",
"type": "factual",
"verified": true,
"evidence": "Counted 12 fields in field_info.json"
}
],
"user_notes_summary": {
"uncertainties": ["Used 2023 data, may be stale"],
"needs_review": [],
"workarounds": ["Fell back to text overlay for non-fillable fields"]
},
"eval_feedback": {
"suggestions": [
{
"assertion": "The output includes the name 'John Smith'",
"reason": "A hallucinated document that mentions the name would also pass"
}
],
"overall": "Assertions check presence but not correctness."
}
}
```
**Fields:**
- `expectations[]`: Graded expectations with evidence
- `summary`: Aggregate pass/fail counts
- `execution_metrics`: Tool usage and output size (from executor's metrics.json)
- `timing`: Wall clock timing (from timing.json)
- `claims`: Extracted and verified claims from the output
- `user_notes_summary`: Issues flagged by the executor
- `eval_feedback`: (optional) Improvement suggestions for the evals, only present when the grader identifies issues worth raising
---
## metrics.json
Output from the executor agent. Located at `<run-dir>/outputs/metrics.json`.
```json
{
"tool_calls": {
"Read": 5,
"Write": 2,
"Bash": 8,
"Edit": 1,
"Glob": 2,
"Grep": 0
},
"total_tool_calls": 18,
"total_steps": 6,
"files_created": ["filled_form.pdf", "field_values.json"],
"errors_encountered": 0,
"output_chars": 12450,
"transcript_chars": 3200
}
```
**Fields:**
- `tool_calls`: Count per tool type
- `total_tool_calls`: Sum of all tool calls
- `total_steps`: Number of major execution steps
- `files_created`: List of output files created
- `errors_encountered`: Number of errors during execution
- `output_chars`: Total character count of output files
- `transcript_chars`: Character count of transcript
---
## timing.json
Wall clock timing for a run. Located at `<run-dir>/timing.json`.
**How to capture:** When a subagent task completes, the task notification includes `total_tokens` and `duration_ms`. Save these immediately — they are not persisted anywhere else and cannot be recovered after the fact.
```json
{
"total_tokens": 84852,
"duration_ms": 23332,
"total_duration_seconds": 23.3,
"executor_start": "2026-01-15T10:30:00Z",
"executor_end": "2026-01-15T10:32:45Z",
"executor_duration_seconds": 165.0,
"grader_start": "2026-01-15T10:32:46Z",
"grader_end": "2026-01-15T10:33:12Z",
"grader_duration_seconds": 26.0
}
```
---
## benchmark.json
Output from Benchmark mode. Located at `benchmarks/<timestamp>/benchmark.json`.
```json
{
"metadata": {
"skill_name": "pdf",
"skill_path": "/path/to/pdf",
"executor_model": "claude-sonnet-4-20250514",
"analyzer_model": "most-capable-model",
"timestamp": "2026-01-15T10:30:00Z",
"evals_run": [1, 2, 3],
"runs_per_configuration": 3
},
"runs": [
{
"eval_id": 1,
"eval_name": "Ocean",
"configuration": "with_skill",
"run_number": 1,
"result": {
"pass_rate": 0.85,
"passed": 6,
"failed": 1,
"total": 7,
"time_seconds": 42.5,
"tokens": 3800,
"tool_calls": 18,
"errors": 0
},
"expectations": [
{"text": "...", "passed": true, "evidence": "..."}
],
"notes": [
"Used 2023 data, may be stale",
"Fell back to text overlay for non-fillable fields"
]
}
],
"run_summary": {
"with_skill": {
"pass_rate": {"mean": 0.85, "stddev": 0.05, "min": 0.80, "max": 0.90},
"time_seconds": {"mean": 45.0, "stddev": 12.0, "min": 32.0, "max": 58.0},
"tokens": {"mean": 3800, "stddev": 400, "min": 3200, "max": 4100}
},
"without_skill": {
"pass_rate": {"mean": 0.35, "stddev": 0.08, "min": 0.28, "max": 0.45},
"time_seconds": {"mean": 32.0, "stddev": 8.0, "min": 24.0, "max": 42.0},
"tokens": {"mean": 2100, "stddev": 300, "min": 1800, "max": 2500}
},
"delta": {
"pass_rate": "+0.50",
"time_seconds": "+13.0",
"tokens": "+1700"
}
},
"notes": [
"Assertion 'Output is a PDF file' passes 100% in both configurations - may not differentiate skill value",
"Eval 3 shows high variance (50% ± 40%) - may be flaky or model-dependent",
"Without-skill runs consistently fail on table extraction expectations",
"Skill adds 13s average execution time but improves pass rate by 50%"
]
}
```
**Fields:**
- `metadata`: Information about the benchmark run
- `skill_name`: Name of the skill
- `timestamp`: When the benchmark was run
- `evals_run`: List of eval names or IDs
- `runs_per_configuration`: Number of runs per config (e.g. 3)
- `runs[]`: Individual run results
- `eval_id`: Numeric eval identifier
- `eval_name`: Human-readable eval name (used as section header in the viewer)
- `configuration`: Must be `"with_skill"` or `"without_skill"` (the viewer uses this exact string for grouping and color coding)
- `run_number`: Integer run number (1, 2, 3...)
- `result`: Nested object with `pass_rate`, `passed`, `total`, `time_seconds`, `tokens`, `errors`
- `run_summary`: Statistical aggregates per configuration
- `with_skill` / `without_skill`: Each contains `pass_rate`, `time_seconds`, `tokens` objects with `mean` and `stddev` fields
- `delta`: Difference strings like `"+0.50"`, `"+13.0"`, `"+1700"`
- `notes`: Freeform observations from the analyzer
**Important:** The viewer reads these field names exactly. Using `config` instead of `configuration`, or putting `pass_rate` at the top level of a run instead of nested under `result`, will cause the viewer to show empty/zero values. Always reference this schema when generating benchmark.json manually.
---
## comparison.json
Output from blind comparator. Located at `<grading-dir>/comparison-N.json`.
```json
{
"winner": "A",
"reasoning": "Output A provides a complete solution with proper formatting and all required fields. Output B is missing the date field and has formatting inconsistencies.",
"rubric": {
"A": {
"content": {
"correctness": 5,
"completeness": 5,
"accuracy": 4
},
"structure": {
"organization": 4,
"formatting": 5,
"usability": 4
},
"content_score": 4.7,
"structure_score": 4.3,
"overall_score": 9.0
},
"B": {
"content": {
"correctness": 3,
"completeness": 2,
"accuracy": 3
},
"structure": {
"organization": 3,
"formatting": 2,
"usability": 3
},
"content_score": 2.7,
"structure_score": 2.7,
"overall_score": 5.4
}
},
"output_quality": {
"A": {
"score": 9,
"strengths": ["Complete solution", "Well-formatted", "All fields present"],
"weaknesses": ["Minor style inconsistency in header"]
},
"B": {
"score": 5,
"strengths": ["Readable output", "Correct basic structure"],
"weaknesses": ["Missing date field", "Formatting inconsistencies", "Partial data extraction"]
}
},
"expectation_results": {
"A": {
"passed": 4,
"total": 5,
"pass_rate": 0.80,
"details": [
{"text": "Output includes name", "passed": true}
]
},
"B": {
"passed": 3,
"total": 5,
"pass_rate": 0.60,
"details": [
{"text": "Output includes name", "passed": true}
]
}
}
}
```
---
## analysis.json
Output from post-hoc analyzer. Located at `<grading-dir>/analysis.json`.
```json
{
"comparison_summary": {
"winner": "A",
"winner_skill": "path/to/winner/skill",
"loser_skill": "path/to/loser/skill",
"comparator_reasoning": "Brief summary of why comparator chose winner"
},
"winner_strengths": [
"Clear step-by-step instructions for handling multi-page documents",
"Included validation script that caught formatting errors"
],
"loser_weaknesses": [
"Vague instruction 'process the document appropriately' led to inconsistent behavior",
"No script for validation, agent had to improvise"
],
"instruction_following": {
"winner": {
"score": 9,
"issues": ["Minor: skipped optional logging step"]
},
"loser": {
"score": 6,
"issues": [
"Did not use the skill's formatting template",
"Invented own approach instead of following step 3"
]
}
},
"improvement_suggestions": [
{
"priority": "high",
"category": "instructions",
"suggestion": "Replace 'process the document appropriately' with explicit steps",
"expected_impact": "Would eliminate ambiguity that caused inconsistent behavior"
}
],
"transcript_insights": {
"winner_execution_pattern": "Read skill -> Followed 5-step process -> Used validation script",
"loser_execution_pattern": "Read skill -> Unclear on approach -> Tried 3 different methods"
}
}
```

0
dot_claude/skills/skill-creator/scripts/empty___init__.py Normal file
View File

401
dot_claude/skills/skill-creator/scripts/executable_aggregate_benchmark.py Executable file
View File

@@ -0,0 +1,401 @@
#!/usr/bin/env python3
"""
Aggregate individual run results into benchmark summary statistics.
Reads grading.json files from run directories and produces:
- run_summary with mean, stddev, min, max for each metric
- delta between with_skill and without_skill configurations
Usage:
python aggregate_benchmark.py <benchmark_dir>
Example:
python aggregate_benchmark.py benchmarks/2026-01-15T10-30-00/
The script supports two directory layouts:
Workspace layout (from skill-creator iterations):
<benchmark_dir>/
└── eval-N/
├── with_skill/
│ ├── run-1/grading.json
│ └── run-2/grading.json
└── without_skill/
├── run-1/grading.json
└── run-2/grading.json
Legacy layout (with runs/ subdirectory):
<benchmark_dir>/
└── runs/
└── eval-N/
├── with_skill/
│ └── run-1/grading.json
└── without_skill/
└── run-1/grading.json
"""
import argparse
import json
import math
import sys
from datetime import datetime, timezone
from pathlib import Path
def calculate_stats(values: list[float]) -> dict:
"""Calculate mean, stddev, min, max for a list of values."""
if not values:
return {"mean": 0.0, "stddev": 0.0, "min": 0.0, "max": 0.0}
n = len(values)
mean = sum(values) / n
if n > 1:
variance = sum((x - mean) ** 2 for x in values) / (n - 1)
stddev = math.sqrt(variance)
else:
stddev = 0.0
return {
"mean": round(mean, 4),
"stddev": round(stddev, 4),
"min": round(min(values), 4),
"max": round(max(values), 4)
}
def load_run_results(benchmark_dir: Path) -> dict:
"""
Load all run results from a benchmark directory.
Returns dict keyed by config name (e.g. "with_skill"/"without_skill",
or "new_skill"/"old_skill"), each containing a list of run results.
"""
# Support both layouts: eval dirs directly under benchmark_dir, or under runs/
runs_dir = benchmark_dir / "runs"
if runs_dir.exists():
search_dir = runs_dir
elif list(benchmark_dir.glob("eval-*")):
search_dir = benchmark_dir
else:
print(f"No eval directories found in {benchmark_dir} or {benchmark_dir / 'runs'}")
return {}
results: dict[str, list] = {}
for eval_idx, eval_dir in enumerate(sorted(search_dir.glob("eval-*"))):
metadata_path = eval_dir / "eval_metadata.json"
if metadata_path.exists():
try:
with open(metadata_path) as mf:
eval_id = json.load(mf).get("eval_id", eval_idx)
except (json.JSONDecodeError, OSError):
eval_id = eval_idx
else:
try:
eval_id = int(eval_dir.name.split("-")[1])
except ValueError:
eval_id = eval_idx
# Discover config directories dynamically rather than hardcoding names
for config_dir in sorted(eval_dir.iterdir()):
if not config_dir.is_dir():
continue
# Skip non-config directories (inputs, outputs, etc.)
if not list(config_dir.glob("run-*")):
continue
config = config_dir.name
if config not in results:
results[config] = []
for run_dir in sorted(config_dir.glob("run-*")):
run_number = int(run_dir.name.split("-")[1])
grading_file = run_dir / "grading.json"
if not grading_file.exists():
print(f"Warning: grading.json not found in {run_dir}")
continue
try:
with open(grading_file) as f:
grading = json.load(f)
except json.JSONDecodeError as e:
print(f"Warning: Invalid JSON in {grading_file}: {e}")
continue
# Extract metrics
result = {
"eval_id": eval_id,
"run_number": run_number,
"pass_rate": grading.get("summary", {}).get("pass_rate", 0.0),
"passed": grading.get("summary", {}).get("passed", 0),
"failed": grading.get("summary", {}).get("failed", 0),
"total": grading.get("summary", {}).get("total", 0),
}
# Extract timing — check grading.json first, then sibling timing.json
timing = grading.get("timing", {})
result["time_seconds"] = timing.get("total_duration_seconds", 0.0)
timing_file = run_dir / "timing.json"
if result["time_seconds"] == 0.0 and timing_file.exists():
try:
with open(timing_file) as tf:
timing_data = json.load(tf)
result["time_seconds"] = timing_data.get("total_duration_seconds", 0.0)
result["tokens"] = timing_data.get("total_tokens", 0)
except json.JSONDecodeError:
pass
# Extract metrics if available
metrics = grading.get("execution_metrics", {})
result["tool_calls"] = metrics.get("total_tool_calls", 0)
if not result.get("tokens"):
result["tokens"] = metrics.get("output_chars", 0)
result["errors"] = metrics.get("errors_encountered", 0)
# Extract expectations — viewer requires fields: text, passed, evidence
raw_expectations = grading.get("expectations", [])
for exp in raw_expectations:
if "text" not in exp or "passed" not in exp:
print(f"Warning: expectation in {grading_file} missing required fields (text, passed, evidence): {exp}")
result["expectations"] = raw_expectations
# Extract notes from user_notes_summary
notes_summary = grading.get("user_notes_summary", {})
notes = []
notes.extend(notes_summary.get("uncertainties", []))
notes.extend(notes_summary.get("needs_review", []))
notes.extend(notes_summary.get("workarounds", []))
result["notes"] = notes
results[config].append(result)
return results
def aggregate_results(results: dict) -> dict:
"""
Aggregate run results into summary statistics.
Returns run_summary with stats for each configuration and delta.
"""
run_summary = {}
configs = list(results.keys())
for config in configs:
runs = results.get(config, [])
if not runs:
run_summary[config] = {
"pass_rate": {"mean": 0.0, "stddev": 0.0, "min": 0.0, "max": 0.0},
"time_seconds": {"mean": 0.0, "stddev": 0.0, "min": 0.0, "max": 0.0},
"tokens": {"mean": 0, "stddev": 0, "min": 0, "max": 0}
}
continue
pass_rates = [r["pass_rate"] for r in runs]
times = [r["time_seconds"] for r in runs]
tokens = [r.get("tokens", 0) for r in runs]
run_summary[config] = {
"pass_rate": calculate_stats(pass_rates),
"time_seconds": calculate_stats(times),
"tokens": calculate_stats(tokens)
}
# Calculate delta between the first two configs (if two exist)
if len(configs) >= 2:
primary = run_summary.get(configs[0], {})
baseline = run_summary.get(configs[1], {})
else:
primary = run_summary.get(configs[0], {}) if configs else {}
baseline = {}
delta_pass_rate = primary.get("pass_rate", {}).get("mean", 0) - baseline.get("pass_rate", {}).get("mean", 0)
delta_time = primary.get("time_seconds", {}).get("mean", 0) - baseline.get("time_seconds", {}).get("mean", 0)
delta_tokens = primary.get("tokens", {}).get("mean", 0) - baseline.get("tokens", {}).get("mean", 0)
run_summary["delta"] = {
"pass_rate": f"{delta_pass_rate:+.2f}",
"time_seconds": f"{delta_time:+.1f}",
"tokens": f"{delta_tokens:+.0f}"
}
return run_summary
def generate_benchmark(benchmark_dir: Path, skill_name: str = "", skill_path: str = "") -> dict:
"""
Generate complete benchmark.json from run results.
"""
results = load_run_results(benchmark_dir)
run_summary = aggregate_results(results)
# Build runs array for benchmark.json
runs = []
for config in results:
for result in results[config]:
runs.append({
"eval_id": result["eval_id"],
"configuration": config,
"run_number": result["run_number"],
"result": {
"pass_rate": result["pass_rate"],
"passed": result["passed"],
"failed": result["failed"],
"total": result["total"],
"time_seconds": result["time_seconds"],
"tokens": result.get("tokens", 0),
"tool_calls": result.get("tool_calls", 0),
"errors": result.get("errors", 0)
},
"expectations": result["expectations"],
"notes": result["notes"]
})
# Determine eval IDs from results
eval_ids = sorted(set(
r["eval_id"]
for config in results.values()
for r in config
))
benchmark = {
"metadata": {
"skill_name": skill_name or "<skill-name>",
"skill_path": skill_path or "<path/to/skill>",
"executor_model": "<model-name>",
"analyzer_model": "<model-name>",
"timestamp": datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ"),
"evals_run": eval_ids,
"runs_per_configuration": 3
},
"runs": runs,
"run_summary": run_summary,
"notes": [] # To be filled by analyzer
}
return benchmark
def generate_markdown(benchmark: dict) -> str:
"""Generate human-readable benchmark.md from benchmark data."""
metadata = benchmark["metadata"]
run_summary = benchmark["run_summary"]
# Determine config names (excluding "delta")
configs = [k for k in run_summary if k != "delta"]
config_a = configs[0] if len(configs) >= 1 else "config_a"
config_b = configs[1] if len(configs) >= 2 else "config_b"
label_a = config_a.replace("_", " ").title()
label_b = config_b.replace("_", " ").title()
lines = [
f"# Skill Benchmark: {metadata['skill_name']}",
"",
f"**Model**: {metadata['executor_model']}",
f"**Date**: {metadata['timestamp']}",
f"**Evals**: {', '.join(map(str, metadata['evals_run']))} ({metadata['runs_per_configuration']} runs each per configuration)",
"",
"## Summary",
"",
f"| Metric | {label_a} | {label_b} | Delta |",
"|--------|------------|---------------|-------|",
]
a_summary = run_summary.get(config_a, {})
b_summary = run_summary.get(config_b, {})
delta = run_summary.get("delta", {})
# Format pass rate
a_pr = a_summary.get("pass_rate", {})
b_pr = b_summary.get("pass_rate", {})
lines.append(f"| Pass Rate | {a_pr.get('mean', 0)*100:.0f}% ± {a_pr.get('stddev', 0)*100:.0f}% | {b_pr.get('mean', 0)*100:.0f}% ± {b_pr.get('stddev', 0)*100:.0f}% | {delta.get('pass_rate', '—')} |")
# Format time
a_time = a_summary.get("time_seconds", {})
b_time = b_summary.get("time_seconds", {})
lines.append(f"| Time | {a_time.get('mean', 0):.1f}s ± {a_time.get('stddev', 0):.1f}s | {b_time.get('mean', 0):.1f}s ± {b_time.get('stddev', 0):.1f}s | {delta.get('time_seconds', '—')}s |")
# Format tokens
a_tokens = a_summary.get("tokens", {})
b_tokens = b_summary.get("tokens", {})
lines.append(f"| Tokens | {a_tokens.get('mean', 0):.0f} ± {a_tokens.get('stddev', 0):.0f} | {b_tokens.get('mean', 0):.0f} ± {b_tokens.get('stddev', 0):.0f} | {delta.get('tokens', '—')} |")
# Notes section
if benchmark.get("notes"):
lines.extend([
"",
"## Notes",
""
])
for note in benchmark["notes"]:
lines.append(f"- {note}")
return "\n".join(lines)
def main():
parser = argparse.ArgumentParser(
description="Aggregate benchmark run results into summary statistics"
)
parser.add_argument(
"benchmark_dir",
type=Path,
help="Path to the benchmark directory"
)
parser.add_argument(
"--skill-name",
default="",
help="Name of the skill being benchmarked"
)
parser.add_argument(
"--skill-path",
default="",
help="Path to the skill being benchmarked"
)
parser.add_argument(
"--output", "-o",
type=Path,
help="Output path for benchmark.json (default: <benchmark_dir>/benchmark.json)"
)
args = parser.parse_args()
if not args.benchmark_dir.exists():
print(f"Directory not found: {args.benchmark_dir}")
sys.exit(1)
# Generate benchmark
benchmark = generate_benchmark(args.benchmark_dir, args.skill_name, args.skill_path)
# Determine output paths
output_json = args.output or (args.benchmark_dir / "benchmark.json")
output_md = output_json.with_suffix(".md")
# Write benchmark.json
with open(output_json, "w") as f:
json.dump(benchmark, f, indent=2)
print(f"Generated: {output_json}")
# Write benchmark.md
markdown = generate_markdown(benchmark)
with open(output_md, "w") as f:
f.write(markdown)
print(f"Generated: {output_md}")
# Print summary
run_summary = benchmark["run_summary"]
configs = [k for k in run_summary if k != "delta"]
delta = run_summary.get("delta", {})
print(f"\nSummary:")
for config in configs:
pr = run_summary[config]["pass_rate"]["mean"]
label = config.replace("_", " ").title()
print(f" {label}: {pr*100:.1f}% pass rate")
print(f" Delta: {delta.get('pass_rate', '—')}")
if __name__ == "__main__":
main()

326
dot_claude/skills/skill-creator/scripts/executable_generate_report.py Executable file
View File

@@ -0,0 +1,326 @@
#!/usr/bin/env python3
"""Generate an HTML report from run_loop.py output.
Takes the JSON output from run_loop.py and generates a visual HTML report
showing each description attempt with check/x for each test case.
Distinguishes between train and test queries.
"""
import argparse
import html
import json
import sys
from pathlib import Path
def generate_html(data: dict, auto_refresh: bool = False, skill_name: str = "") -> str:
"""Generate HTML report from loop output data. If auto_refresh is True, adds a meta refresh tag."""
history = data.get("history", [])
holdout = data.get("holdout", 0)
title_prefix = html.escape(skill_name + " \u2014 ") if skill_name else ""
# Get all unique queries from train and test sets, with should_trigger info
train_queries: list[dict] = []
test_queries: list[dict] = []
if history:
for r in history[0].get("train_results", history[0].get("results", [])):
train_queries.append({"query": r["query"], "should_trigger": r.get("should_trigger", True)})
if history[0].get("test_results"):
for r in history[0].get("test_results", []):
test_queries.append({"query": r["query"], "should_trigger": r.get("should_trigger", True)})
refresh_tag = ' <meta http-equiv="refresh" content="5">\n' if auto_refresh else ""
html_parts = ["""<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
""" + refresh_tag + """ <title>""" + title_prefix + """Skill Description Optimization</title>
<link rel="preconnect" href="https://fonts.googleapis.com">
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
<link href="https://fonts.googleapis.com/css2?family=Poppins:wght@500;600&family=Lora:wght@400;500&display=swap" rel="stylesheet">
<style>
body {
font-family: 'Lora', Georgia, serif;
max-width: 100%;
margin: 0 auto;
padding: 20px;
background: #faf9f5;
color: #141413;
}
h1 { font-family: 'Poppins', sans-serif; color: #141413; }
.explainer {
background: white;
padding: 15px;
border-radius: 6px;
margin-bottom: 20px;
border: 1px solid #e8e6dc;
color: #b0aea5;
font-size: 0.875rem;
line-height: 1.6;
}
.summary {
background: white;
padding: 15px;
border-radius: 6px;
margin-bottom: 20px;
border: 1px solid #e8e6dc;
}
.summary p { margin: 5px 0; }
.best { color: #788c5d; font-weight: bold; }
.table-container {
overflow-x: auto;
width: 100%;
}
table {
border-collapse: collapse;
background: white;
border: 1px solid #e8e6dc;
border-radius: 6px;
font-size: 12px;
min-width: 100%;
}
th, td {
padding: 8px;
text-align: left;
border: 1px solid #e8e6dc;
white-space: normal;
word-wrap: break-word;
}
th {
font-family: 'Poppins', sans-serif;
background: #141413;
color: #faf9f5;
font-weight: 500;
}
th.test-col {
background: #6a9bcc;
}
th.query-col { min-width: 200px; }
td.description {
font-family: monospace;
font-size: 11px;
word-wrap: break-word;
max-width: 400px;
}
td.result {
text-align: center;
font-size: 16px;
min-width: 40px;
}
td.test-result {
background: #f0f6fc;
}
.pass { color: #788c5d; }
.fail { color: #c44; }
.rate {
font-size: 9px;
color: #b0aea5;
display: block;
}
tr:hover { background: #faf9f5; }
.score {
display: inline-block;
padding: 2px 6px;
border-radius: 4px;
font-weight: bold;
font-size: 11px;
}
.score-good { background: #eef2e8; color: #788c5d; }
.score-ok { background: #fef3c7; color: #d97706; }
.score-bad { background: #fceaea; color: #c44; }
.train-label { color: #b0aea5; font-size: 10px; }
.test-label { color: #6a9bcc; font-size: 10px; font-weight: bold; }
.best-row { background: #f5f8f2; }
th.positive-col { border-bottom: 3px solid #788c5d; }
th.negative-col { border-bottom: 3px solid #c44; }
th.test-col.positive-col { border-bottom: 3px solid #788c5d; }
th.test-col.negative-col { border-bottom: 3px solid #c44; }
.legend { font-family: 'Poppins', sans-serif; display: flex; gap: 20px; margin-bottom: 10px; font-size: 13px; align-items: center; }
.legend-item { display: flex; align-items: center; gap: 6px; }
.legend-swatch { width: 16px; height: 16px; border-radius: 3px; display: inline-block; }
.swatch-positive { background: #141413; border-bottom: 3px solid #788c5d; }
.swatch-negative { background: #141413; border-bottom: 3px solid #c44; }
.swatch-test { background: #6a9bcc; }
.swatch-train { background: #141413; }
</style>
</head>
<body>
<h1>""" + title_prefix + """Skill Description Optimization</h1>
<div class="explainer">
<strong>Optimizing your skill's description.</strong> This page updates automatically as Claude tests different versions of your skill's description. Each row is an iteration — a new description attempt. The columns show test queries: green checkmarks mean the skill triggered correctly (or correctly didn't trigger), red crosses mean it got it wrong. The "Train" score shows performance on queries used to improve the description; the "Test" score shows performance on held-out queries the optimizer hasn't seen. When it's done, Claude will apply the best-performing description to your skill.
</div>
"""]
# Summary section
best_test_score = data.get('best_test_score')
best_train_score = data.get('best_train_score')
html_parts.append(f"""
<div class="summary">
<p><strong>Original:</strong> {html.escape(data.get('original_description', 'N/A'))}</p>
<p class="best"><strong>Best:</strong> {html.escape(data.get('best_description', 'N/A'))}</p>
<p><strong>Best Score:</strong> {data.get('best_score', 'N/A')} {'(test)' if best_test_score else '(train)'}</p>
<p><strong>Iterations:</strong> {data.get('iterations_run', 0)} | <strong>Train:</strong> {data.get('train_size', '?')} | <strong>Test:</strong> {data.get('test_size', '?')}</p>
</div>
""")
# Legend
html_parts.append("""
<div class="legend">
<span style="font-weight:600">Query columns:</span>
<span class="legend-item"><span class="legend-swatch swatch-positive"></span> Should trigger</span>
<span class="legend-item"><span class="legend-swatch swatch-negative"></span> Should NOT trigger</span>
<span class="legend-item"><span class="legend-swatch swatch-train"></span> Train</span>
<span class="legend-item"><span class="legend-swatch swatch-test"></span> Test</span>
</div>
""")
# Table header
html_parts.append("""
<div class="table-container">
<table>
<thead>
<tr>
<th>Iter</th>
<th>Train</th>
<th>Test</th>
<th class="query-col">Description</th>
""")
# Add column headers for train queries
for qinfo in train_queries:
polarity = "positive-col" if qinfo["should_trigger"] else "negative-col"
html_parts.append(f' <th class="{polarity}">{html.escape(qinfo["query"])}</th>\n')
# Add column headers for test queries (different color)
for qinfo in test_queries:
polarity = "positive-col" if qinfo["should_trigger"] else "negative-col"
html_parts.append(f' <th class="test-col {polarity}">{html.escape(qinfo["query"])}</th>\n')
html_parts.append(""" </tr>
</thead>
<tbody>
""")
# Find best iteration for highlighting
if test_queries:
best_iter = max(history, key=lambda h: h.get("test_passed") or 0).get("iteration")
else:
best_iter = max(history, key=lambda h: h.get("train_passed", h.get("passed", 0))).get("iteration")
# Add rows for each iteration
for h in history:
iteration = h.get("iteration", "?")
train_passed = h.get("train_passed", h.get("passed", 0))
train_total = h.get("train_total", h.get("total", 0))
test_passed = h.get("test_passed")
test_total = h.get("test_total")
description = h.get("description", "")
train_results = h.get("train_results", h.get("results", []))
test_results = h.get("test_results", [])
# Create lookups for results by query
train_by_query = {r["query"]: r for r in train_results}
test_by_query = {r["query"]: r for r in test_results} if test_results else {}
# Compute aggregate correct/total runs across all retries
def aggregate_runs(results: list[dict]) -> tuple[int, int]:
correct = 0
total = 0
for r in results:
runs = r.get("runs", 0)
triggers = r.get("triggers", 0)
total += runs
if r.get("should_trigger", True):
correct += triggers
else:
correct += runs - triggers
return correct, total
train_correct, train_runs = aggregate_runs(train_results)
test_correct, test_runs = aggregate_runs(test_results)
# Determine score classes
def score_class(correct: int, total: int) -> str:
if total > 0:
ratio = correct / total
if ratio >= 0.8:
return "score-good"
elif ratio >= 0.5:
return "score-ok"
return "score-bad"
train_class = score_class(train_correct, train_runs)
test_class = score_class(test_correct, test_runs)
row_class = "best-row" if iteration == best_iter else ""
html_parts.append(f""" <tr class="{row_class}">
<td>{iteration}</td>
<td><span class="score {train_class}">{train_correct}/{train_runs}</span></td>
<td><span class="score {test_class}">{test_correct}/{test_runs}</span></td>
<td class="description">{html.escape(description)}</td>
""")
# Add result for each train query
for qinfo in train_queries:
r = train_by_query.get(qinfo["query"], {})
did_pass = r.get("pass", False)
triggers = r.get("triggers", 0)
runs = r.get("runs", 0)
icon = "✓" if did_pass else "✗"
css_class = "pass" if did_pass else "fail"
html_parts.append(f' <td class="result {css_class}">{icon}<span class="rate">{triggers}/{runs}</span></td>\n')
# Add result for each test query (with different background)
for qinfo in test_queries:
r = test_by_query.get(qinfo["query"], {})
did_pass = r.get("pass", False)
triggers = r.get("triggers", 0)
runs = r.get("runs", 0)
icon = "✓" if did_pass else "✗"
css_class = "pass" if did_pass else "fail"
html_parts.append(f' <td class="result test-result {css_class}">{icon}<span class="rate">{triggers}/{runs}</span></td>\n')
html_parts.append(" </tr>\n")
html_parts.append(""" </tbody>
</table>
</div>
""")
html_parts.append("""
</body>
</html>
""")
return "".join(html_parts)
def main():
parser = argparse.ArgumentParser(description="Generate HTML report from run_loop output")
parser.add_argument("input", help="Path to JSON output from run_loop.py (or - for stdin)")
parser.add_argument("-o", "--output", default=None, help="Output HTML file (default: stdout)")
parser.add_argument("--skill-name", default="", help="Skill name to include in the report title")
args = parser.parse_args()
if args.input == "-":
data = json.load(sys.stdin)
else:
data = json.loads(Path(args.input).read_text())
html_output = generate_html(data, skill_name=args.skill_name)
if args.output:
Path(args.output).write_text(html_output)
print(f"Report written to {args.output}", file=sys.stderr)
else:
print(html_output)
if __name__ == "__main__":
main()

247
dot_claude/skills/skill-creator/scripts/executable_improve_description.py Executable file
View File

@@ -0,0 +1,247 @@
#!/usr/bin/env python3
"""Improve a skill description based on eval results.
Takes eval results (from run_eval.py) and generates an improved description
by calling `claude -p` as a subprocess (same auth pattern as run_eval.py —
uses the session's Claude Code auth, no separate ANTHROPIC_API_KEY needed).
"""
import argparse
import json
import os
import re
import subprocess
import sys
from pathlib import Path
from scripts.utils import parse_skill_md
def _call_claude(prompt: str, model: str | None, timeout: int = 300) -> str:
"""Run `claude -p` with the prompt on stdin and return the text response.
Prompt goes over stdin (not argv) because it embeds the full SKILL.md
body and can easily exceed comfortable argv length.
"""
cmd = ["claude", "-p", "--output-format", "text"]
if model:
cmd.extend(["--model", model])
# Remove CLAUDECODE env var to allow nesting claude -p inside a
# Claude Code session. The guard is for interactive terminal conflicts;
# programmatic subprocess usage is safe. Same pattern as run_eval.py.
env = {k: v for k, v in os.environ.items() if k != "CLAUDECODE"}
result = subprocess.run(
cmd,
input=prompt,
capture_output=True,
text=True,
env=env,
timeout=timeout,
)
if result.returncode != 0:
raise RuntimeError(
f"claude -p exited {result.returncode}\nstderr: {result.stderr}"
)
return result.stdout
def improve_description(
skill_name: str,
skill_content: str,
current_description: str,
eval_results: dict,
history: list[dict],
model: str,
test_results: dict | None = None,
log_dir: Path | None = None,
iteration: int | None = None,
) -> str:
"""Call Claude to improve the description based on eval results."""
failed_triggers = [
r for r in eval_results["results"]
if r["should_trigger"] and not r["pass"]
]
false_triggers = [
r for r in eval_results["results"]
if not r["should_trigger"] and not r["pass"]
]
# Build scores summary
train_score = f"{eval_results['summary']['passed']}/{eval_results['summary']['total']}"
if test_results:
test_score = f"{test_results['summary']['passed']}/{test_results['summary']['total']}"
scores_summary = f"Train: {train_score}, Test: {test_score}"
else:
scores_summary = f"Train: {train_score}"
prompt = f"""You are optimizing a skill description for a Claude Code skill called "{skill_name}". A "skill" is sort of like a prompt, but with progressive disclosure -- there's a title and description that Claude sees when deciding whether to use the skill, and then if it does use the skill, it reads the .md file which has lots more details and potentially links to other resources in the skill folder like helper files and scripts and additional documentation or examples.
The description appears in Claude's "available_skills" list. When a user sends a query, Claude decides whether to invoke the skill based solely on the title and on this description. Your goal is to write a description that triggers for relevant queries, and doesn't trigger for irrelevant ones.
Here's the current description:
<current_description>
"{current_description}"
</current_description>
Current scores ({scores_summary}):
<scores_summary>
"""
if failed_triggers:
prompt += "FAILED TO TRIGGER (should have triggered but didn't):\n"
for r in failed_triggers:
prompt += f' - "{r["query"]}" (triggered {r["triggers"]}/{r["runs"]} times)\n'
prompt += "\n"
if false_triggers:
prompt += "FALSE TRIGGERS (triggered but shouldn't have):\n"
for r in false_triggers:
prompt += f' - "{r["query"]}" (triggered {r["triggers"]}/{r["runs"]} times)\n'
prompt += "\n"
if history:
prompt += "PREVIOUS ATTEMPTS (do NOT repeat these — try something structurally different):\n\n"
for h in history:
train_s = f"{h.get('train_passed', h.get('passed', 0))}/{h.get('train_total', h.get('total', 0))}"
test_s = f"{h.get('test_passed', '?')}/{h.get('test_total', '?')}" if h.get('test_passed') is not None else None
score_str = f"train={train_s}" + (f", test={test_s}" if test_s else "")
prompt += f'<attempt {score_str}>\n'
prompt += f'Description: "{h["description"]}"\n'
if "results" in h:
prompt += "Train results:\n"
for r in h["results"]:
status = "PASS" if r["pass"] else "FAIL"
prompt += f' [{status}] "{r["query"][:80]}" (triggered {r["triggers"]}/{r["runs"]})\n'
if h.get("note"):
prompt += f'Note: {h["note"]}\n'
prompt += "</attempt>\n\n"
prompt += f"""</scores_summary>
Skill content (for context on what the skill does):
<skill_content>
{skill_content}
</skill_content>
Based on the failures, write a new and improved description that is more likely to trigger correctly. When I say "based on the failures", it's a bit of a tricky line to walk because we don't want to overfit to the specific cases you're seeing. So what I DON'T want you to do is produce an ever-expanding list of specific queries that this skill should or shouldn't trigger for. Instead, try to generalize from the failures to broader categories of user intent and situations where this skill would be useful or not useful. The reason for this is twofold:
1. Avoid overfitting
2. The list might get loooong and it's injected into ALL queries and there might be a lot of skills, so we don't want to blow too much space on any given description.
Concretely, your description should not be more than about 100-200 words, even if that comes at the cost of accuracy. There is a hard limit of 1024 characters — descriptions over that will be truncated, so stay comfortably under it.
Here are some tips that we've found to work well in writing these descriptions:
- The skill should be phrased in the imperative -- "Use this skill for" rather than "this skill does"
- The skill description should focus on the user's intent, what they are trying to achieve, vs. the implementation details of how the skill works.
- The description competes with other skills for Claude's attention — make it distinctive and immediately recognizable.
- If you're getting lots of failures after repeated attempts, change things up. Try different sentence structures or wordings.
I'd encourage you to be creative and mix up the style in different iterations since you'll have multiple opportunities to try different approaches and we'll just grab the highest-scoring one at the end.
Please respond with only the new description text in <new_description> tags, nothing else."""
text = _call_claude(prompt, model)
match = re.search(r"<new_description>(.*?)</new_description>", text, re.DOTALL)
description = match.group(1).strip().strip('"') if match else text.strip().strip('"')
transcript: dict = {
"iteration": iteration,
"prompt": prompt,
"response": text,
"parsed_description": description,
"char_count": len(description),
"over_limit": len(description) > 1024,
}
# Safety net: the prompt already states the 1024-char hard limit, but if
# the model blew past it anyway, make one fresh single-turn call that
# quotes the too-long version and asks for a shorter rewrite. (The old
# SDK path did this as a true multi-turn; `claude -p` is one-shot, so we
# inline the prior output into the new prompt instead.)
if len(description) > 1024:
shorten_prompt = (
f"{prompt}\n\n"
f"---\n\n"
f"A previous attempt produced this description, which at "
f"{len(description)} characters is over the 1024-character hard limit:\n\n"
f'"{description}"\n\n'
f"Rewrite it to be under 1024 characters while keeping the most "
f"important trigger words and intent coverage. Respond with only "
f"the new description in <new_description> tags."
)
shorten_text = _call_claude(shorten_prompt, model)
match = re.search(r"<new_description>(.*?)</new_description>", shorten_text, re.DOTALL)
shortened = match.group(1).strip().strip('"') if match else shorten_text.strip().strip('"')
transcript["rewrite_prompt"] = shorten_prompt
transcript["rewrite_response"] = shorten_text
transcript["rewrite_description"] = shortened
transcript["rewrite_char_count"] = len(shortened)
description = shortened
transcript["final_description"] = description
if log_dir:
log_dir.mkdir(parents=True, exist_ok=True)
log_file = log_dir / f"improve_iter_{iteration or 'unknown'}.json"
log_file.write_text(json.dumps(transcript, indent=2))
return description
def main():
parser = argparse.ArgumentParser(description="Improve a skill description based on eval results")
parser.add_argument("--eval-results", required=True, help="Path to eval results JSON (from run_eval.py)")
parser.add_argument("--skill-path", required=True, help="Path to skill directory")
parser.add_argument("--history", default=None, help="Path to history JSON (previous attempts)")
parser.add_argument("--model", required=True, help="Model for improvement")
parser.add_argument("--verbose", action="store_true", help="Print thinking to stderr")
args = parser.parse_args()
skill_path = Path(args.skill_path)
if not (skill_path / "SKILL.md").exists():
print(f"Error: No SKILL.md found at {skill_path}", file=sys.stderr)
sys.exit(1)
eval_results = json.loads(Path(args.eval_results).read_text())
history = []
if args.history:
history = json.loads(Path(args.history).read_text())
name, _, content = parse_skill_md(skill_path)
current_description = eval_results["description"]
if args.verbose:
print(f"Current: {current_description}", file=sys.stderr)
print(f"Score: {eval_results['summary']['passed']}/{eval_results['summary']['total']}", file=sys.stderr)
new_description = improve_description(
skill_name=name,
skill_content=content,
current_description=current_description,
eval_results=eval_results,
history=history,
model=args.model,
)
if args.verbose:
print(f"Improved: {new_description}", file=sys.stderr)
# Output as JSON with both the new description and updated history
output = {
"description": new_description,
"history": history + [{
"description": current_description,
"passed": eval_results["summary"]["passed"],
"failed": eval_results["summary"]["failed"],
"total": eval_results["summary"]["total"],
"results": eval_results["results"],
}],
}
print(json.dumps(output, indent=2))
if __name__ == "__main__":
main()

310
dot_claude/skills/skill-creator/scripts/executable_literal_run_eval.py Executable file
View File

@@ -0,0 +1,310 @@
#!/usr/bin/env python3
"""Run trigger evaluation for a skill description.
Tests whether a skill's description causes Claude to trigger (read the skill)
for a set of queries. Outputs results as JSON.
"""
import argparse
import json
import os
import select
import subprocess
import sys
import time
import uuid
from concurrent.futures import ProcessPoolExecutor, as_completed
from pathlib import Path
from scripts.utils import parse_skill_md
def find_project_root() -> Path:
"""Find the project root by walking up from cwd looking for .claude/.
Mimics how Claude Code discovers its project root, so the command file
we create ends up where claude -p will look for it.
"""
current = Path.cwd()
for parent in [current, *current.parents]:
if (parent / ".claude").is_dir():
return parent
return current
def run_single_query(
query: str,
skill_name: str,
skill_description: str,
timeout: int,
project_root: str,
model: str | None = None,
) -> bool:
"""Run a single query and return whether the skill was triggered.
Creates a command file in .claude/commands/ so it appears in Claude's
available_skills list, then runs `claude -p` with the raw query.
Uses --include-partial-messages to detect triggering early from
stream events (content_block_start) rather than waiting for the
full assistant message, which only arrives after tool execution.
"""
unique_id = uuid.uuid4().hex[:8]
clean_name = f"{skill_name}-skill-{unique_id}"
project_commands_dir = Path(project_root) / ".claude" / "commands"
command_file = project_commands_dir / f"{clean_name}.md"
try:
project_commands_dir.mkdir(parents=True, exist_ok=True)
# Use YAML block scalar to avoid breaking on quotes in description
indented_desc = "\n ".join(skill_description.split("\n"))
command_content = (
f"---\n"
f"description: |\n"
f" {indented_desc}\n"
f"---\n\n"
f"# {skill_name}\n\n"
f"This skill handles: {skill_description}\n"
)
command_file.write_text(command_content)
cmd = [
"claude",
"-p", query,
"--output-format", "stream-json",
"--verbose",
"--include-partial-messages",
]
if model:
cmd.extend(["--model", model])
# Remove CLAUDECODE env var to allow nesting claude -p inside a
# Claude Code session. The guard is for interactive terminal conflicts;
# programmatic subprocess usage is safe.
env = {k: v for k, v in os.environ.items() if k != "CLAUDECODE"}
process = subprocess.Popen(
cmd,
stdout=subprocess.PIPE,
stderr=subprocess.DEVNULL,
cwd=project_root,
env=env,
)
triggered = False
start_time = time.time()
buffer = ""
# Track state for stream event detection
pending_tool_name = None
accumulated_json = ""
try:
while time.time() - start_time < timeout:
if process.poll() is not None:
remaining = process.stdout.read()
if remaining:
buffer += remaining.decode("utf-8", errors="replace")
break
ready, _, _ = select.select([process.stdout], [], [], 1.0)
if not ready:
continue
chunk = os.read(process.stdout.fileno(), 8192)
if not chunk:
break
buffer += chunk.decode("utf-8", errors="replace")
while "\n" in buffer:
line, buffer = buffer.split("\n", 1)
line = line.strip()
if not line:
continue
try:
event = json.loads(line)
except json.JSONDecodeError:
continue
# Early detection via stream events
if event.get("type") == "stream_event":
se = event.get("event", {})
se_type = se.get("type", "")
if se_type == "content_block_start":
cb = se.get("content_block", {})
if cb.get("type") == "tool_use":
tool_name = cb.get("name", "")
if tool_name in ("Skill", "Read"):
pending_tool_name = tool_name
accumulated_json = ""
else:
return False
elif se_type == "content_block_delta" and pending_tool_name:
delta = se.get("delta", {})
if delta.get("type") == "input_json_delta":
accumulated_json += delta.get("partial_json", "")
if clean_name in accumulated_json:
return True
elif se_type in ("content_block_stop", "message_stop"):
if pending_tool_name:
return clean_name in accumulated_json
if se_type == "message_stop":
return False
# Fallback: full assistant message
elif event.get("type") == "assistant":
message = event.get("message", {})
for content_item in message.get("content", []):
if content_item.get("type") != "tool_use":
continue
tool_name = content_item.get("name", "")
tool_input = content_item.get("input", {})
if tool_name == "Skill" and clean_name in tool_input.get("skill", ""):
triggered = True
elif tool_name == "Read" and clean_name in tool_input.get("file_path", ""):
triggered = True
return triggered
elif event.get("type") == "result":
return triggered
finally:
# Clean up process on any exit path (return, exception, timeout)
if process.poll() is None:
process.kill()
process.wait()
return triggered
finally:
if command_file.exists():
command_file.unlink()
def run_eval(
eval_set: list[dict],
skill_name: str,
description: str,
num_workers: int,
timeout: int,
project_root: Path,
runs_per_query: int = 1,
trigger_threshold: float = 0.5,
model: str | None = None,
) -> dict:
"""Run the full eval set and return results."""
results = []
with ProcessPoolExecutor(max_workers=num_workers) as executor:
future_to_info = {}
for item in eval_set:
for run_idx in range(runs_per_query):
future = executor.submit(
run_single_query,
item["query"],
skill_name,
description,
timeout,
str(project_root),
model,
)
future_to_info[future] = (item, run_idx)
query_triggers: dict[str, list[bool]] = {}
query_items: dict[str, dict] = {}
for future in as_completed(future_to_info):
item, _ = future_to_info[future]
query = item["query"]
query_items[query] = item
if query not in query_triggers:
query_triggers[query] = []
try:
query_triggers[query].append(future.result())
except Exception as e:
print(f"Warning: query failed: {e}", file=sys.stderr)
query_triggers[query].append(False)
for query, triggers in query_triggers.items():
item = query_items[query]
trigger_rate = sum(triggers) / len(triggers)
should_trigger = item["should_trigger"]
if should_trigger:
did_pass = trigger_rate >= trigger_threshold
else:
did_pass = trigger_rate < trigger_threshold
results.append({
"query": query,
"should_trigger": should_trigger,
"trigger_rate": trigger_rate,
"triggers": sum(triggers),
"runs": len(triggers),
"pass": did_pass,
})
passed = sum(1 for r in results if r["pass"])
total = len(results)
return {
"skill_name": skill_name,
"description": description,
"results": results,
"summary": {
"total": total,
"passed": passed,
"failed": total - passed,
},
}
def main():
parser = argparse.ArgumentParser(description="Run trigger evaluation for a skill description")
parser.add_argument("--eval-set", required=True, help="Path to eval set JSON file")
parser.add_argument("--skill-path", required=True, help="Path to skill directory")
parser.add_argument("--description", default=None, help="Override description to test")
parser.add_argument("--num-workers", type=int, default=10, help="Number of parallel workers")
parser.add_argument("--timeout", type=int, default=30, help="Timeout per query in seconds")
parser.add_argument("--runs-per-query", type=int, default=3, help="Number of runs per query")
parser.add_argument("--trigger-threshold", type=float, default=0.5, help="Trigger rate threshold")
parser.add_argument("--model", default=None, help="Model to use for claude -p (default: user's configured model)")
parser.add_argument("--verbose", action="store_true", help="Print progress to stderr")
args = parser.parse_args()
eval_set = json.loads(Path(args.eval_set).read_text())
skill_path = Path(args.skill_path)
if not (skill_path / "SKILL.md").exists():
print(f"Error: No SKILL.md found at {skill_path}", file=sys.stderr)
sys.exit(1)
name, original_description, content = parse_skill_md(skill_path)
description = args.description or original_description
project_root = find_project_root()
if args.verbose:
print(f"Evaluating: {description}", file=sys.stderr)
output = run_eval(
eval_set=eval_set,
skill_name=name,
description=description,
num_workers=args.num_workers,
timeout=args.timeout,
project_root=project_root,
runs_per_query=args.runs_per_query,
trigger_threshold=args.trigger_threshold,
model=args.model,
)
if args.verbose:
summary = output["summary"]
print(f"Results: {summary['passed']}/{summary['total']} passed", file=sys.stderr)
for r in output["results"]:
status = "PASS" if r["pass"] else "FAIL"
rate_str = f"{r['triggers']}/{r['runs']}"
print(f" [{status}] rate={rate_str} expected={r['should_trigger']}: {r['query'][:70]}", file=sys.stderr)
print(json.dumps(output, indent=2))
if __name__ == "__main__":
main()

328
dot_claude/skills/skill-creator/scripts/executable_literal_run_loop.py Executable file
View File

@@ -0,0 +1,328 @@
#!/usr/bin/env python3
"""Run the eval + improve loop until all pass or max iterations reached.
Combines run_eval.py and improve_description.py in a loop, tracking history
and returning the best description found. Supports train/test split to prevent
overfitting.
"""
import argparse
import json
import random
import sys
import tempfile
import time
import webbrowser
from pathlib import Path
from scripts.generate_report import generate_html
from scripts.improve_description import improve_description
from scripts.run_eval import find_project_root, run_eval
from scripts.utils import parse_skill_md
def split_eval_set(eval_set: list[dict], holdout: float, seed: int = 42) -> tuple[list[dict], list[dict]]:
"""Split eval set into train and test sets, stratified by should_trigger."""
random.seed(seed)
# Separate by should_trigger
trigger = [e for e in eval_set if e["should_trigger"]]
no_trigger = [e for e in eval_set if not e["should_trigger"]]
# Shuffle each group
random.shuffle(trigger)
random.shuffle(no_trigger)
# Calculate split points
n_trigger_test = max(1, int(len(trigger) * holdout))
n_no_trigger_test = max(1, int(len(no_trigger) * holdout))
# Split
test_set = trigger[:n_trigger_test] + no_trigger[:n_no_trigger_test]
train_set = trigger[n_trigger_test:] + no_trigger[n_no_trigger_test:]
return train_set, test_set
def run_loop(
eval_set: list[dict],
skill_path: Path,
description_override: str | None,
num_workers: int,
timeout: int,
max_iterations: int,
runs_per_query: int,
trigger_threshold: float,
holdout: float,
model: str,
verbose: bool,
live_report_path: Path | None = None,
log_dir: Path | None = None,
) -> dict:
"""Run the eval + improvement loop."""
project_root = find_project_root()
name, original_description, content = parse_skill_md(skill_path)
current_description = description_override or original_description
# Split into train/test if holdout > 0
if holdout > 0:
train_set, test_set = split_eval_set(eval_set, holdout)
if verbose:
print(f"Split: {len(train_set)} train, {len(test_set)} test (holdout={holdout})", file=sys.stderr)
else:
train_set = eval_set
test_set = []
history = []
exit_reason = "unknown"
for iteration in range(1, max_iterations + 1):
if verbose:
print(f"\n{'='*60}", file=sys.stderr)
print(f"Iteration {iteration}/{max_iterations}", file=sys.stderr)
print(f"Description: {current_description}", file=sys.stderr)
print(f"{'='*60}", file=sys.stderr)
# Evaluate train + test together in one batch for parallelism
all_queries = train_set + test_set
t0 = time.time()
all_results = run_eval(
eval_set=all_queries,
skill_name=name,
description=current_description,
num_workers=num_workers,
timeout=timeout,
project_root=project_root,
runs_per_query=runs_per_query,
trigger_threshold=trigger_threshold,
model=model,
)
eval_elapsed = time.time() - t0
# Split results back into train/test by matching queries
train_queries_set = {q["query"] for q in train_set}
train_result_list = [r for r in all_results["results"] if r["query"] in train_queries_set]
test_result_list = [r for r in all_results["results"] if r["query"] not in train_queries_set]
train_passed = sum(1 for r in train_result_list if r["pass"])
train_total = len(train_result_list)
train_summary = {"passed": train_passed, "failed": train_total - train_passed, "total": train_total}
train_results = {"results": train_result_list, "summary": train_summary}
if test_set:
test_passed = sum(1 for r in test_result_list if r["pass"])
test_total = len(test_result_list)
test_summary = {"passed": test_passed, "failed": test_total - test_passed, "total": test_total}
test_results = {"results": test_result_list, "summary": test_summary}
else:
test_results = None
test_summary = None
history.append({
"iteration": iteration,
"description": current_description,
"train_passed": train_summary["passed"],
"train_failed": train_summary["failed"],
"train_total": train_summary["total"],
"train_results": train_results["results"],
"test_passed": test_summary["passed"] if test_summary else None,
"test_failed": test_summary["failed"] if test_summary else None,
"test_total": test_summary["total"] if test_summary else None,
"test_results": test_results["results"] if test_results else None,
# For backward compat with report generator
"passed": train_summary["passed"],
"failed": train_summary["failed"],
"total": train_summary["total"],
"results": train_results["results"],
})
# Write live report if path provided
if live_report_path:
partial_output = {
"original_description": original_description,
"best_description": current_description,
"best_score": "in progress",
"iterations_run": len(history),
"holdout": holdout,
"train_size": len(train_set),
"test_size": len(test_set),
"history": history,
}
live_report_path.write_text(generate_html(partial_output, auto_refresh=True, skill_name=name))
if verbose:
def print_eval_stats(label, results, elapsed):
pos = [r for r in results if r["should_trigger"]]
neg = [r for r in results if not r["should_trigger"]]
tp = sum(r["triggers"] for r in pos)
pos_runs = sum(r["runs"] for r in pos)
fn = pos_runs - tp
fp = sum(r["triggers"] for r in neg)
neg_runs = sum(r["runs"] for r in neg)
tn = neg_runs - fp
total = tp + tn + fp + fn
precision = tp / (tp + fp) if (tp + fp) > 0 else 1.0
recall = tp / (tp + fn) if (tp + fn) > 0 else 1.0
accuracy = (tp + tn) / total if total > 0 else 0.0
print(f"{label}: {tp+tn}/{total} correct, precision={precision:.0%} recall={recall:.0%} accuracy={accuracy:.0%} ({elapsed:.1f}s)", file=sys.stderr)
for r in results:
status = "PASS" if r["pass"] else "FAIL"
rate_str = f"{r['triggers']}/{r['runs']}"
print(f" [{status}] rate={rate_str} expected={r['should_trigger']}: {r['query'][:60]}", file=sys.stderr)
print_eval_stats("Train", train_results["results"], eval_elapsed)
if test_summary:
print_eval_stats("Test ", test_results["results"], 0)
if train_summary["failed"] == 0:
exit_reason = f"all_passed (iteration {iteration})"
if verbose:
print(f"\nAll train queries passed on iteration {iteration}!", file=sys.stderr)
break
if iteration == max_iterations:
exit_reason = f"max_iterations ({max_iterations})"
if verbose:
print(f"\nMax iterations reached ({max_iterations}).", file=sys.stderr)
break
# Improve the description based on train results
if verbose:
print(f"\nImproving description...", file=sys.stderr)
t0 = time.time()
# Strip test scores from history so improvement model can't see them
blinded_history = [
{k: v for k, v in h.items() if not k.startswith("test_")}
for h in history
]
new_description = improve_description(
skill_name=name,
skill_content=content,
current_description=current_description,
eval_results=train_results,
history=blinded_history,
model=model,
log_dir=log_dir,
iteration=iteration,
)
improve_elapsed = time.time() - t0
if verbose:
print(f"Proposed ({improve_elapsed:.1f}s): {new_description}", file=sys.stderr)
current_description = new_description
# Find the best iteration by TEST score (or train if no test set)
if test_set:
best = max(history, key=lambda h: h["test_passed"] or 0)
best_score = f"{best['test_passed']}/{best['test_total']}"
else:
best = max(history, key=lambda h: h["train_passed"])
best_score = f"{best['train_passed']}/{best['train_total']}"
if verbose:
print(f"\nExit reason: {exit_reason}", file=sys.stderr)
print(f"Best score: {best_score} (iteration {best['iteration']})", file=sys.stderr)
return {
"exit_reason": exit_reason,
"original_description": original_description,
"best_description": best["description"],
"best_score": best_score,
"best_train_score": f"{best['train_passed']}/{best['train_total']}",
"best_test_score": f"{best['test_passed']}/{best['test_total']}" if test_set else None,
"final_description": current_description,
"iterations_run": len(history),
"holdout": holdout,
"train_size": len(train_set),
"test_size": len(test_set),
"history": history,
}
def main():
parser = argparse.ArgumentParser(description="Run eval + improve loop")
parser.add_argument("--eval-set", required=True, help="Path to eval set JSON file")
parser.add_argument("--skill-path", required=True, help="Path to skill directory")
parser.add_argument("--description", default=None, help="Override starting description")
parser.add_argument("--num-workers", type=int, default=10, help="Number of parallel workers")
parser.add_argument("--timeout", type=int, default=30, help="Timeout per query in seconds")
parser.add_argument("--max-iterations", type=int, default=5, help="Max improvement iterations")
parser.add_argument("--runs-per-query", type=int, default=3, help="Number of runs per query")
parser.add_argument("--trigger-threshold", type=float, default=0.5, help="Trigger rate threshold")
parser.add_argument("--holdout", type=float, default=0.4, help="Fraction of eval set to hold out for testing (0 to disable)")
parser.add_argument("--model", required=True, help="Model for improvement")
parser.add_argument("--verbose", action="store_true", help="Print progress to stderr")
parser.add_argument("--report", default="auto", help="Generate HTML report at this path (default: 'auto' for temp file, 'none' to disable)")
parser.add_argument("--results-dir", default=None, help="Save all outputs (results.json, report.html, log.txt) to a timestamped subdirectory here")
args = parser.parse_args()
eval_set = json.loads(Path(args.eval_set).read_text())
skill_path = Path(args.skill_path)
if not (skill_path / "SKILL.md").exists():
print(f"Error: No SKILL.md found at {skill_path}", file=sys.stderr)
sys.exit(1)
name, _, _ = parse_skill_md(skill_path)
# Set up live report path
if args.report != "none":
if args.report == "auto":
timestamp = time.strftime("%Y%m%d_%H%M%S")
live_report_path = Path(tempfile.gettempdir()) / f"skill_description_report_{skill_path.name}_{timestamp}.html"
else:
live_report_path = Path(args.report)
# Open the report immediately so the user can watch
live_report_path.write_text("<html><body><h1>Starting optimization loop...</h1><meta http-equiv='refresh' content='5'></body></html>")
webbrowser.open(str(live_report_path))
else:
live_report_path = None
# Determine output directory (create before run_loop so logs can be written)
if args.results_dir:
timestamp = time.strftime("%Y-%m-%d_%H%M%S")
results_dir = Path(args.results_dir) / timestamp
results_dir.mkdir(parents=True, exist_ok=True)
else:
results_dir = None
log_dir = results_dir / "logs" if results_dir else None
output = run_loop(
eval_set=eval_set,
skill_path=skill_path,
description_override=args.description,
num_workers=args.num_workers,
timeout=args.timeout,
max_iterations=args.max_iterations,
runs_per_query=args.runs_per_query,
trigger_threshold=args.trigger_threshold,
holdout=args.holdout,
model=args.model,
verbose=args.verbose,
live_report_path=live_report_path,
log_dir=log_dir,
)
# Save JSON output
json_output = json.dumps(output, indent=2)
print(json_output)
if results_dir:
(results_dir / "results.json").write_text(json_output)
# Write final HTML report (without auto-refresh)
if live_report_path:
live_report_path.write_text(generate_html(output, auto_refresh=False, skill_name=name))
print(f"\nReport: {live_report_path}", file=sys.stderr)
if results_dir and live_report_path:
(results_dir / "report.html").write_text(generate_html(output, auto_refresh=False, skill_name=name))
if results_dir:
print(f"Results saved to: {results_dir}", file=sys.stderr)
if __name__ == "__main__":
main()

136
dot_claude/skills/skill-creator/scripts/executable_package_skill.py Executable file
View File

@@ -0,0 +1,136 @@
#!/usr/bin/env python3
"""
Skill Packager - Creates a distributable .skill file of a skill folder
Usage:
python utils/package_skill.py <path/to/skill-folder> [output-directory]
Example:
python utils/package_skill.py skills/public/my-skill
python utils/package_skill.py skills/public/my-skill ./dist
"""
import fnmatch
import sys
import zipfile
from pathlib import Path
from scripts.quick_validate import validate_skill
# Patterns to exclude when packaging skills.
EXCLUDE_DIRS = {"__pycache__", "node_modules"}
EXCLUDE_GLOBS = {"*.pyc"}
EXCLUDE_FILES = {".DS_Store"}
# Directories excluded only at the skill root (not when nested deeper).
ROOT_EXCLUDE_DIRS = {"evals"}
def should_exclude(rel_path: Path) -> bool:
"""Check if a path should be excluded from packaging."""
parts = rel_path.parts
if any(part in EXCLUDE_DIRS for part in parts):
return True
# rel_path is relative to skill_path.parent, so parts[0] is the skill
# folder name and parts[1] (if present) is the first subdir.
if len(parts) > 1 and parts[1] in ROOT_EXCLUDE_DIRS:
return True
name = rel_path.name
if name in EXCLUDE_FILES:
return True
return any(fnmatch.fnmatch(name, pat) for pat in EXCLUDE_GLOBS)
def package_skill(skill_path, output_dir=None):
"""
Package a skill folder into a .skill file.
Args:
skill_path: Path to the skill folder
output_dir: Optional output directory for the .skill file (defaults to current directory)
Returns:
Path to the created .skill file, or None if error
"""
skill_path = Path(skill_path).resolve()
# Validate skill folder exists
if not skill_path.exists():
print(f"❌ Error: Skill folder not found: {skill_path}")
return None
if not skill_path.is_dir():
print(f"❌ Error: Path is not a directory: {skill_path}")
return None
# Validate SKILL.md exists
skill_md = skill_path / "SKILL.md"
if not skill_md.exists():
print(f"❌ Error: SKILL.md not found in {skill_path}")
return None
# Run validation before packaging
print("🔍 Validating skill...")
valid, message = validate_skill(skill_path)
if not valid:
print(f"❌ Validation failed: {message}")
print(" Please fix the validation errors before packaging.")
return None
print(f"✅ {message}\n")
# Determine output location
skill_name = skill_path.name
if output_dir:
output_path = Path(output_dir).resolve()
output_path.mkdir(parents=True, exist_ok=True)
else:
output_path = Path.cwd()
skill_filename = output_path / f"{skill_name}.skill"
# Create the .skill file (zip format)
try:
with zipfile.ZipFile(skill_filename, 'w', zipfile.ZIP_DEFLATED) as zipf:
# Walk through the skill directory, excluding build artifacts
for file_path in skill_path.rglob('*'):
if not file_path.is_file():
continue
arcname = file_path.relative_to(skill_path.parent)
if should_exclude(arcname):
print(f" Skipped: {arcname}")
continue
zipf.write(file_path, arcname)
print(f" Added: {arcname}")
print(f"\n✅ Successfully packaged skill to: {skill_filename}")
return skill_filename
except Exception as e:
print(f"❌ Error creating .skill file: {e}")
return None
def main():
if len(sys.argv) < 2:
print("Usage: python utils/package_skill.py <path/to/skill-folder> [output-directory]")
print("\nExample:")
print(" python utils/package_skill.py skills/public/my-skill")
print(" python utils/package_skill.py skills/public/my-skill ./dist")
sys.exit(1)
skill_path = sys.argv[1]
output_dir = sys.argv[2] if len(sys.argv) > 2 else None
print(f"📦 Packaging skill: {skill_path}")
if output_dir:
print(f" Output directory: {output_dir}")
print()
result = package_skill(skill_path, output_dir)
if result:
sys.exit(0)
else:
sys.exit(1)
if __name__ == "__main__":
main()

103
dot_claude/skills/skill-creator/scripts/executable_quick_validate.py Executable file
View File

@@ -0,0 +1,103 @@
#!/usr/bin/env python3
"""
Quick validation script for skills - minimal version
"""
import sys
import os
import re
import yaml
from pathlib import Path
def validate_skill(skill_path):
"""Basic validation of a skill"""
skill_path = Path(skill_path)
# Check SKILL.md exists
skill_md = skill_path / 'SKILL.md'
if not skill_md.exists():
return False, "SKILL.md not found"
# Read and validate frontmatter
content = skill_md.read_text()
if not content.startswith('---'):
return False, "No YAML frontmatter found"
# Extract frontmatter
match = re.match(r'^---\n(.*?)\n---', content, re.DOTALL)
if not match:
return False, "Invalid frontmatter format"
frontmatter_text = match.group(1)
# Parse YAML frontmatter
try:
frontmatter = yaml.safe_load(frontmatter_text)
if not isinstance(frontmatter, dict):
return False, "Frontmatter must be a YAML dictionary"
except yaml.YAMLError as e:
return False, f"Invalid YAML in frontmatter: {e}"
# Define allowed properties
ALLOWED_PROPERTIES = {'name', 'description', 'license', 'allowed-tools', 'metadata', 'compatibility'}
# Check for unexpected properties (excluding nested keys under metadata)
unexpected_keys = set(frontmatter.keys()) - ALLOWED_PROPERTIES
if unexpected_keys:
return False, (
f"Unexpected key(s) in SKILL.md frontmatter: {', '.join(sorted(unexpected_keys))}. "
f"Allowed properties are: {', '.join(sorted(ALLOWED_PROPERTIES))}"
)
# Check required fields
if 'name' not in frontmatter:
return False, "Missing 'name' in frontmatter"
if 'description' not in frontmatter:
return False, "Missing 'description' in frontmatter"
# Extract name for validation
name = frontmatter.get('name', '')
if not isinstance(name, str):
return False, f"Name must be a string, got {type(name).__name__}"
name = name.strip()
if name:
# Check naming convention (kebab-case: lowercase with hyphens)
if not re.match(r'^[a-z0-9-]+$', name):
return False, f"Name '{name}' should be kebab-case (lowercase letters, digits, and hyphens only)"
if name.startswith('-') or name.endswith('-') or '--' in name:
return False, f"Name '{name}' cannot start/end with hyphen or contain consecutive hyphens"
# Check name length (max 64 characters per spec)
if len(name) > 64:
return False, f"Name is too long ({len(name)} characters). Maximum is 64 characters."
# Extract and validate description
description = frontmatter.get('description', '')
if not isinstance(description, str):
return False, f"Description must be a string, got {type(description).__name__}"
description = description.strip()
if description:
# Check for angle brackets
if '<' in description or '>' in description:
return False, "Description cannot contain angle brackets (< or >)"
# Check description length (max 1024 characters per spec)
if len(description) > 1024:
return False, f"Description is too long ({len(description)} characters). Maximum is 1024 characters."
# Validate compatibility field if present (optional)
compatibility = frontmatter.get('compatibility', '')
if compatibility:
if not isinstance(compatibility, str):
return False, f"Compatibility must be a string, got {type(compatibility).__name__}"
if len(compatibility) > 500:
return False, f"Compatibility is too long ({len(compatibility)} characters). Maximum is 500 characters."
return True, "Skill is valid!"
if __name__ == "__main__":
if len(sys.argv) != 2:
print("Usage: python quick_validate.py <skill_directory>")
sys.exit(1)
valid, message = validate_skill(sys.argv[1])
print(message)
sys.exit(0 if valid else 1)

47
dot_claude/skills/skill-creator/scripts/utils.py Normal file
View File

@@ -0,0 +1,47 @@
"""Shared utilities for skill-creator scripts."""
from pathlib import Path
def parse_skill_md(skill_path: Path) -> tuple[str, str, str]:
"""Parse a SKILL.md file, returning (name, description, full_content)."""
content = (skill_path / "SKILL.md").read_text()
lines = content.split("\n")
if lines[0].strip() != "---":
raise ValueError("SKILL.md missing frontmatter (no opening ---)")
end_idx = None
for i, line in enumerate(lines[1:], start=1):
if line.strip() == "---":
end_idx = i
break
if end_idx is None:
raise ValueError("SKILL.md missing frontmatter (no closing ---)")
name = ""
description = ""
frontmatter_lines = lines[1:end_idx]
i = 0
while i < len(frontmatter_lines):
line = frontmatter_lines[i]
if line.startswith("name:"):
name = line[len("name:"):].strip().strip('"').strip("'")
elif line.startswith("description:"):
value = line[len("description:"):].strip()
# Handle YAML multiline indicators (>, |, >-, |-)
if value in (">", "|", ">-", "|-"):
continuation_lines: list[str] = []
i += 1
while i < len(frontmatter_lines) and (frontmatter_lines[i].startswith(" ") or frontmatter_lines[i].startswith("\t")):
continuation_lines.append(frontmatter_lines[i].strip())
i += 1
description = " ".join(continuation_lines)
continue
else:
description = value.strip('"').strip("'")
i += 1
return name, description, content