claude-skill-spec-audit
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data by scanning
SKILL.mdfiles located in the user's local skill directories (~/.claude/skills/and.claude/skills/). A malicious skill could include instructions designed to hijack the agent's behavior during the audit process. - Ingestion points: The skill glob-searches and reads
SKILL.mdfiles from multiple local directories (SKILL.md). - Boundary markers: Absent. The instructions do not specify the use of delimiters or provide warnings to ignore embedded instructions when parsing the content of external skills.
- Capability inventory: The skill uses the
Edittool to modify local files and analyzes other skills for high-privilege operations such as filesystem writes, package installations (brew install), and remote operations (git push,Slack). - Sanitization: Absent. There is no mention of sanitizing or escaping the content of the audited skills before the agent processes them for validation.
Audit Metadata