caveman-stats
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface for indirect prompt injection by processing session logs and outputting data to the local file system.\n
- Ingestion points: The skill reads directly from session logs which contain model outputs and untrusted user inputs (SKILL.md).\n
- Boundary markers: There are no explicit delimiters or instructions provided to separate the untrusted session log content from the structured wiki report template.\n
- Capability inventory: The skill has instructions to create and modify markdown files in the llmwiki/ directory, representing a file system write capability (SKILL.md).\n
- Sanitization: No sanitization or validation of the ingested session data is mentioned before it is processed and written to the wiki logs.
Audit Metadata