| name | skill-comply |
| description | Automated compliance measurement -- generate behavioral specs from skill/rule files, run scenarios at decreasing prompt strictness, capture tool call traces, and report compliance scores. |
| origin | ECC |
Skill Comply -- Automated Compliance Measurement
Measures whether coding agents actually follow skills, rules, or agent definitions.
When to Use
- After adding new rules or skills, to verify agent compliance
- Periodically as part of quality maintenance
- When asking "is this rule actually being followed?"
How It Works
- Auto-generate specs: Extract expected behavioral sequences from any .md file
- Auto-generate scenarios: Create test prompts with decreasing strictness (supportive -> neutral -> competing)
- Run agent: Execute
claude -p and capture tool call traces via stream-json
- Classify: Match tool calls against spec steps using LLM (not regex)
- Check ordering: Verify temporal ordering deterministically
- Report: Generate self-contained reports with spec, prompts, and timelines
Supported Targets
- Skills (
skills/*/SKILL.md): Workflow skills like search-first, TDD guides
- Rules (
rules/common/*.md): Mandatory rules like testing.md, security.md
- Agent definitions (
agents/*.md): Whether an agent gets invoked when expected
Key Concept: Prompt Independence
Measures whether a skill/rule is followed even when the prompt does not explicitly support it.
Usage
uv run python -m scripts.run ~/.claude/rules/common/testing.md
uv run python -m scripts.run --dry-run ~/.claude/skills/search-first/SKILL.md
Report Contents
- Expected behavioral sequence (auto-generated spec)
- Scenario prompts at each strictness level
- Compliance scores per scenario
- Tool call timelines with classification labels