setup-agent-routing
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is a documentation and routing configuration tool that operates within a restricted scope of local file writes and read-only GitHub inspection.
- [SAFE]: The instructions incorporate explicit security safeguards, such as warning against shell interpolation of tracker data and mandating that environment files be treated as non-executable data.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external metadata from GitHub and local configuration files, creating a potential surface for indirect injection that is mitigated by specific instructions.
- Ingestion points: GitHub issue tracker labels, project data, and repository-level configuration files like .github/agent-loop.env.
- Boundary markers: The skill requires the agent to prepare drafts for user approval and adhere to native repository field conventions.
- Capability inventory: Authorized capabilities are limited to writing markdown documentation to CLAUDE.md, AGENTS.md, and the docs/agents/ directory.
- Sanitization: Explicit instructions prohibit the use of shell interpolation for external text and require strict data-only handling of environment files.
Audit Metadata