claude-skill-spec-audit
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes content from external SKILL.md files located in local directories. This introduces a surface for indirect prompt injection, as malicious instructions within a scanned file could potentially influence the agent's behavior during the audit process. Ingestion points: Reads frontmatter and body from SKILL.md files in ~/.claude/skills/ and .claude/skills/. Boundary markers: None specified to differentiate between the auditor's instructions and the content being audited. Capability inventory: Includes the ability to modify local files via the fix argument. Sanitization: No sanitization or safety-filtering of the ingested file content is defined.
Audit Metadata