ponytail-debt
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute
git grepwith user-provided marker words. The instruction "If the user names their own marker word (/ponytail-debt TODO), grep for that word instead" lacks guidance on sanitizing this input, which could potentially lead to command injection if a malicious search term is provided by the user. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by scanning comments within the codebase. An attacker who can contribute code could embed malicious instructions in these comments that might influence the agent's behavior during the reporting process.
- Ingestion points: File contents (code comments) across the repository via
git grep(SKILL.md). - Boundary markers: None specified to delimit scanned content from agent instructions.
- Capability inventory: The skill performs file system reads and can optionally write a report file named
PONYTAIL-DEBT.md(SKILL.md). - Sanitization: No sanitization or filtering of the scanned comment content is mentioned.
Audit Metadata