perf-investigation-logger
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill captures untrusted user data ("exact user quotes") and persists it directly into the file system, creating a potential surface for instructions that could influence future agent actions. \n
- Ingestion points: User-supplied quotes are ingested as verbatim text according to the instructions in SKILL.md. \n
- Boundary markers: The output format uses basic double-quote delimiters ("") which can be easily bypassed by content containing markdown syntax or instructions. \n
- Capability inventory: The skill directs the agent to perform file-append operations to local files within the {state-dir}/perf/investigations/ directory. \n
- Sanitization: There is no explicit sanitization, filtering, or escaping of the user-provided quotes before they are logged to disk.
Audit Metadata