verify-docs-findable
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes content from external, potentially untrusted repositories to verify documentation effectiveness.
- Ingestion points: The skill reads documentation files and source code from a target repository (referenced as in SKILL.md) to answer specific questions.
- Boundary markers: The sub-agent instructions include the constraint "Answer using ONLY the repo's docs," which serves as a basic boundary marker, although specific delimiters for untrusted content are absent.
- Capability inventory: The sub-agents are restricted to analysis and reporting based on documentation and do not appear to be granted access to sensitive tools such as network operations or file system modification within the prompt template.
- Sanitization: No evidence of sanitization, filtering, or escaping of the repository content was found before the data is passed to the sub-agents for processing.
Audit Metadata