setup-agent-routing
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
ghandgitcommand-line tools to discover repository metadata, such as remote URLs, label lists, and project board configurations. These are read-only operations necessary for the skill's setup purpose. - [DATA_EXPOSURE_AND_EXFILTRATION]: The skill accesses GitHub metadata to populate local documentation templates (e.g.,
CLAUDE.md,docs/agents/). This data is processed locally and not transmitted to external third-party services. - [INDIRECT_PROMPT_INJECTION]: As the skill ingests metadata from GitHub (like label names and project titles), it is theoretically exposed to indirect injection from those sources. However, the skill implements a mandatory human confirmation step, requiring the user to review all findings and drafts before any files are written, which effectively mitigates this risk.
Audit Metadata