agent-loop-triage
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from GitHub issues and pull requests for triage and classification.
- Ingestion points:
SKILL.md(instructions to scan open issues and PRs using theghtool). - Boundary markers: Absent. The instructions do not define delimiters or provide the agent with explicit directives to disregard instructions that may be embedded within the body of the issues or PRs being scanned.
- Capability inventory: The skill utilizes the
ghCLI to read repository data and invokesgh apito fetch user profile information. Malicious content in an issue or PR could attempt to influence the triage classification or the parameters passed to these tools. - Sanitization: Absent. There is no evidence of escaping, validation, or filtering applied to the external content before the agent processes it for classification and metadata generation.
Audit Metadata