ponytail-help
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes features like 'ponytail-debt', 'ponytail-review', and 'ponytail-audit' that ingest data from untrusted sources such as repository files and code comments. This creates an indirect prompt injection surface where malicious instructions could be embedded in the codebase.\n
- Ingestion points: Source code files and 'shortcut:' comments (SKILL.md).\n
- Boundary markers: Absent; the documentation does not define delimiters to separate ingested data from agent instructions.\n
- Capability inventory: The described features require broad read access to the local filesystem and repository.\n
- Sanitization: Absent; no mention of content sanitization or instruction filtering for the ingested data.\n- [SAFE]: The documentation provides standard update instructions for the host environment using official package managers and refers to external documentation on GitHub.
Audit Metadata