ponytail-debt
Pass
Audited by Gen Agent Trust Hub on Aug 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
grepandgit blameon codebase files to find debt markers. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via unsanitized code comments. 1. Ingestion points: Local codebase files accessed via
grep(SKILL.md). 2. Boundary markers: Absent; harvested strings are not enclosed in delimiters. 3. Capability inventory: Shell command execution (grep,git blame) and file writes (SKILL.md). 4. Sanitization: Absent; no filtering or validation is performed on harvested strings.
Audit Metadata