manage-pull-requests
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-generated content from the Harness platform (PR titles, descriptions, and comments), creating an attack surface for instructions embedded in pull request data that could influence agent behavior.
- Ingestion points: Untrusted data enters the agent context via harness_get (pull_request) and harness_list (pr_activity).
- Boundary markers: The instructions do not define delimiters or warnings for the agent to ignore instructions embedded in the retrieved platform content.
- Capability inventory: The skill possesses significant capabilities, including harness_execute (action: merge), harness_delete (pr_comment), and harness_update (pull_request).
- Sanitization: No explicit content validation or sanitization steps are mandated for platform-retrieved data.
Audit Metadata