wiki
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill manages project knowledge by storing and retrieving markdown files in the local
.rcs/wiki/directory, creating a surface for indirect prompt injection. - Ingestion points: Data enters the system via the
contentfield in thewiki_ingestandwiki_addoperations (SKILL.md). - Boundary markers: The instructions lack delimiters or warnings to treat wiki content strictly as data, potentially leading the agent to follow instructions embedded in wiki pages.
- Capability inventory: The skill enables file system write and read operations within the project scope at
.rcs/wiki/(SKILL.md). - Sanitization: No mechanisms for content validation, escaping, or instruction filtering are described for the processed markdown content.
Audit Metadata