no-comments
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through the analysis of untrusted source code. It instructs the agent to interpret and investigate instructions or claims made within code comments, which can be manipulated by an attacker.
- Ingestion points: The skill processes external, untrusted content including user-provided files and branch diffs as specified in the Scope section of SKILL.md.
- Boundary markers: There are no boundary markers or instructions to treat the content of the comments as untrusted data; rather, the agent is told to "Investigate uncertainty" and "trace the named symbol or call" based on the comment's claims.
- Capability inventory: The agent has the capability to write to the filesystem, delete content, and implement root-cause fixes in the code as described in the Act section of SKILL.md.
- Sanitization: The skill does not implement any sanitization or filtering of the comment text before the agent processes it to decide on code modifications.
Audit Metadata