reference-engine
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a mandatory workflow that requires the agent to ingest and act upon data from external, untrusted search results to guide building activities.\n
- Ingestion points: The skill instructs the agent to use
github-searchand documentation searches for "industry leader" patterns in SKILL.md.\n - Boundary markers: There are no instructions to use boundary markers or to ignore potential commands embedded within retrieved reference implementations.\n
- Capability inventory: The agent is directed to "Build: Implement using the reference" (SKILL.md), which involves generating code and architecture based on the external input.\n
- Sanitization: The skill lacks validation or sanitization protocols for the patterns extracted from external search results.\n- [NO_CODE]: The skill is entirely composed of instructional markdown and checklists. It does not contain any executable scripts, binaries, or dynamic code execution components.
Audit Metadata