avibebuilder/claude-prime / .claude/skills/self-evolve
self-evolve
VerifiedThe single config quality skill. Two modes: (1) Default — review/wire pending self-improvement proposals. (2) Audit — analyzes skills, rules, CLAUDE.md, project refs, hooks, and agents against the actual codebase to find inaccuracies, trigger overlaps, stale references, and weak rules, then fixes them. Trigger on 'self-evolve', 'evolve', 'config health check', 'audit config', 'check claude setup', 'apply this rule', 'reorganize rules', 'ok' (when proposals are pending). NOT for improving individual skills (use skill-creator), diagnosing code bugs (use diagnose), or fixing broken code (use fix).
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.4 KB |
| references/audit-guide.md | markdown | 2.6 KB |
| references/quality-dimensions.md | markdown | 1.3 KB |
references/audit-guide.md
View on GitHub# Audit Checklists
## Skills
For each domain skill (skip workflow skills for deep analysis, but DO check their trigger boundaries):
1. **Reference validity**: Extract all file paths from SKILL.md. Glob each — does it exist? Flag broken references as CRITICAL.
2. **Trigger boundaries**: Read this skill's description alongside ALL other descriptions. Look for two skills claiming the same phrasing pattern, or domain skills overlapping with `ask`, `discuss`, `give-plan`, `fix`, or `diagnose`.
3. **Knowledge accuracy**: For factual claims (file paths, tools, architecture), grep to verify. Sample 2-3 claims per skill.
4. **Completeness**: Flag CRITICAL gaps only. Report MODERATE gaps but don't fix them.
## Rules
For each rule file:
1. **Strength**: If it uses "consider", "try to", "generally" — flag as MODERATE. Rules must be imperatives.
2. **Relevance**: Search the codebase for the pattern it protects. If gone → CRITICAL (stale).
3. **Cross-layer overlap**: If a skill already teaches it, the rule is redundant — flag as MODERATE.
## CLAUDE.md
1. **Factual accuracy**: Grep to verify "All X lives in Y" and "We use Z" claims.
2. **Reference validity**: Verify all on-demand reference paths exist.
3. **Length**: Flag if over 200 lines.
4. **Cross-layer consistency**: Flag contradictions between CLAUDE.md, skills, and rules.
## CLAUDE.local.md (if exists)
1. **Separation of concerns**: Flag team-shared instructions that belong in `CLAUDE.md`. Personal file should only contain user-specific content (role, sandbox URLs, preferred test data, workflow quirks).
2. **Conflicts**: Flag contradictions with `CLAUDE.md` or rules — personal prefs must not override team guardrails.
3. **Staleness**: Check for references to paths, URLs, or tools that no longer exist.
## Hooks and settings
1. Verify each hook script path exists and is executable.
2. Check event types are appropriate for the script's purpose.
---
## Report Template
```markdown
## Self-Evolve Report
**Project**: {name}
**Scope**: {scope}
**Pending proposals**: {N or "pipeline not installed"}
### Fixed ({N})
| Component | Issue | Severity | What Changed |
|-----------|-------|----------|--------------|
### Detected but Not Fixed ({N})
| Component | Issue | Severity | Why Not Fixed |
|-----------|-------|----------|---------------|
### Upstream Recommendations (omit when running on claude-prime itself)
| Pattern | Affected Starter/Template | Suggested Fix |
|---------|--------------------------|---------------|
### Health: {verdict}
```
If clean: "Config is healthy. Checked {N} skills, {N} rules, CLAUDE.md ({N} lines), CLAUDE.local.md ({present/absent}), {N} project refs, {N} hooks, {N} agents."
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |