oss-finder
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the GitHub CLI (
gh) to perform repository searches and fetch repository metadata. The instructions include commands such asgh auth status,gh search repos,gh api, andgh repo view. These operations are fundamental to the skill's purpose of discovering and evaluating open-source projects. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data fetched from external GitHub repositories, creating a surface for potential indirect prompt injection attacks.
- Ingestion points: Untrusted data enters the context through
gh search repos(repository descriptions) andgh apicalls (contents ofCONTRIBUTING.mdfiles) inSKILL.md. - Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are present in the instructions when processing external repository data.
- Capability inventory: The skill possesses shell command execution capabilities via the
ghCLI tool. - Sanitization: There is no mention of sanitization, filtering, or escaping of the external content before it is summarized or translated by the agent.
Audit Metadata