warden — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited warden (Agent Skill) and scored it 45/100 (orange). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 1 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 1 flagged
A base64 string of 128+ characters appears in a documentation file. Encoded prompt injection hides the hostile instruction in base64 — invisible to keyword filters — and relies on the agent's ability to decode it at runtime. There is no normal authoring reason to embed a multi-hundred-byte base64 blob in skill docs.
*.sig, SIGNATURES) outside the documentation.Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Run Warden to analyze code changes before committing.
Read the relevant reference when the task requires deeper detail:
| Document | Read When |
|---|---|
${CLAUDE_SKILL_ROOT}/references/cli-reference.md | Full option details, per-command flags, examples |
${CLAUDE_SKILL_ROOT}/references/configuration.md | Editing warden.toml, triggers, patterns, troubleshooting |
${CLAUDE_SKILL_ROOT}/references/config-schema.md | Exact field names, types, and defaults |
${CLAUDE_SKILL_ROOT}/references/creating-skills.md | Writing custom skills, remote skills, skill discovery |
# Analyze uncommitted changes (uses warden.toml triggers)
warden
# Run a specific skill
warden --skill <skill-name>
# Analyze specific files
warden src/auth.ts src/database.ts
# Analyze changes from a git ref
warden main..HEAD
warden HEAD~3
# Auto-apply suggested fixes
warden --fix
# Fail on high-severity findings
warden --fail-on highSet WARDEN_ANTHROPIC_API_KEY or log in via claude login before running.
After making code changes and before committing:
warden to analyze uncommitted changeswarden --fix to auto-apply)Run Warden once to validate work. Do not loop re-running Warden on the same changes.
Severity levels:
high - Must fix before mergemedium - Worth reviewinglow - Minor improvementExit codes: 0 = no findings at or above fail threshold. 1 = findings at or above fail threshold.
Verbosity: -v shows real-time findings. -vv shows debug info (tokens, latency). -q shows errors and summary only.
| Command | Description |
|---|---|
warden | Run analysis (default) |
warden init | Initialize warden.toml and GitHub workflow |
warden add [skill] | Add skill trigger to warden.toml |
warden sync [remote] | Update cached remote skills |
warden setup-app | Create GitHub App via manifest flow |
For full options and flags, read ${CLAUDE_SKILL_ROOT}/references/cli-reference.md.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.