self-audit
Conducts comprehensive health checks on Claude Code, offering quick and full audits to enhance performance and compliance.
Install this skill
Security score
The self-audit skill was audited on Jun 12, 2026 and we found 32 security issues across 1 threat category, including 32 high-severity. Review the findings below before installing.
Categories Tested
Security Issues
Template literal with variable interpolation in command context
| 9 | Before running any checks, execute the bootstrap sequence. Read `${HOME}/.claude/skills/self-audit/bootstrap.md` for platform detection, command mapping, writability check, and first-run initializatio |
Template literal with variable interpolation in command context
| 15 | **0. Self-Audit Integrity.** Run `wc -l ${HOME}/.claude/skills/self-audit/SKILL.md`. If >300 lines → WARNING: self-audit SKILL.md itself exceeds limit. Report as F-SLF-001. |
Template literal with variable interpolation in command context
| 28 | To determine prior audit count, read `${HOME}/.claude/memory/audit-log.md` and count lines matching "| <date> |" format under the "## Pipeline Run Log" section. |
Template literal with variable interpolation in command context
| 34 | Full roster & collaboration modes: `${HOME}/.claude/rules/team.md` |
Template literal with variable interpolation in command context
| 38 | All file paths in this skill use `${HOME}` as the home directory. Resolve `${HOME}` at runtime: |
Template literal with variable interpolation in command context
| 39 | - CLAUDE.md: `${HOME}/.claude/CLAUDE.md` |
Template literal with variable interpolation in command context
| 40 | - Settings: `${HOME}/.claude/settings.json` |
Template literal with variable interpolation in command context
| 41 | - Rules: `${HOME}/.claude/rules/coding.md` |
Template literal with variable interpolation in command context
| 42 | - Memory: `${HOME}/.claude/memory/` |
Template literal with variable interpolation in command context
| 43 | - Skills: `${HOME}/.claude/skills/` |
Template literal with variable interpolation in command context
| 44 | - Audit log: `${HOME}/.claude/memory/audit-log.md` |
Template literal with variable interpolation in command context
| 45 | - Environment: `${HOME}/.claude/memory/environment.md` |
Template literal with variable interpolation in command context
| 46 | - Archive: `${HOME}/.claude/memory/.archive/` |
Template literal with variable interpolation in command context
| 47 | - Yushi agent: `${HOME}/.claude/agents/yushi.md` |
Template literal with variable interpolation in command context
| 48 | - Skill usage inventory: `${HOME}/.claude/memory/skill-usage-inventory.md` |
Template literal with variable interpolation in command context
| 49 | - Memory index: `${HOME}/.claude/memory/MEMORY.md` |
Template literal with variable interpolation in command context
| 50 | - Pitfalls archive: `${HOME}/.claude/rules/pitfalls-archive.md` |
Template literal with variable interpolation in command context
| 51 | - Team rules: `${HOME}/.claude/rules/team.md` |
Template literal with variable interpolation in command context
| 107 | Before counting repetitions, read `${HOME}/.claude/memory/audit-log.md` and check the "## Accepted Exceptions" table. Skip any finding whose ID appears there. |
Template literal with variable interpolation in command context
| 120 | **Quick**: `wc -l ${HOME}/.claude/CLAUDE.md` — line count only. (~12 tokens) |
Template literal with variable interpolation in command context
| 121 | **Full**: Read companion `${HOME}/.claude/skills/self-audit/compress-config.md`. Parse CLAUDE.md sections by `## ` headers. Compute 5-dim health score: D1 Section Balance(25%), D2 Structure Granularit |
Template literal with variable interpolation in command context
| 131 | **Quick**: `ls ${HOME}/.claude/skills/ | wc -l` + `ls ${HOME}/.agents/skills/ | wc -l`. Count only. Do NOT read lockfile. Do NOT check duplicates. (~15 tokens) |
Template literal with variable interpolation in command context
| 136 | **Quick**: `grep -c "sk-" ${HOME}/.claude/settings.json` (plaintext key check) + `grep "skipDangerousModePermissionPrompt" ${HOME}/.claude/settings.json` (dangerous mode check). Do NOT read full file. |
Template literal with variable interpolation in command context
| 141 | **Quick**: `ls ${HOME}/.claude/memory/ | wc -l` (file count). `wc -l ${HOME}/.claude/memory/MEMORY.md` (MEMORY.md size). Do NOT read file contents. (~20 tokens) |
Template literal with variable interpolation in command context
| 146 | **Quick**: `grep "ANTHROPIC_DEFAULT_" ${HOME}/.claude/settings.json` — check if 3 tiers are distinct. Do NOT read full file. (~15 tokens) |
Template literal with variable interpolation in command context
| 156 | **Full only** (items 7-11 are Full mode only). Read `${HOME}/.claude/memory/skill-usage-inventory.md`. Cross-reference installed skills vs invocation count. Flag skills with 0 usage after 30d, flag >2 |
Template literal with variable interpolation in command context
| 165 | **Full only**. List `${HOME}/.claude/memory/.archive/` (archived memory count). Read `${HOME}/.claude/rules/pitfalls-archive.md` (archived pitfalls). Flag archived items recurring in recent conversati |
Template literal with variable interpolation in command context
| 177 | **Quick** (4 atomic checks, ~65 tokens): Run inline checks for (a) disk space <20GB → F-ENV-003, (b) RAM <8GB → F-ENV-006, (c) LibreOffice missing → F-ENV-001, (d) PYTHONIOENCODING unset/GBK → F-ENV-0 |
Template literal with variable interpolation in command context
| 179 | **Full**: Read `${HOME}/.claude/skills/self-audit/environment-checks.md`. Execute: version parsing (8 tools), pip package check (7 packages), network connectivity (read ANTHROPIC_BASE_URL from setting |
Template literal with variable interpolation in command context
| 210 | **Full only**. 24h TTL-gated market scan. Read `${HOME}/.claude/skills/self-audit/market-intelligence.md`. Scan 3 skill marketplace repos via GitHub API (`gh api`), score repos (stars 35% + recency 30 |
Template literal with variable interpolation in command context
| 272 | 1. Append ONE line to `${HOME}/.claude/memory/audit-log.md` under "## Pipeline Run Log": |
Template literal with variable interpolation in command context
| 278 | 2. If environment changed, update `${HOME}/.claude/memory/environment.md` with new snapshot. |