handoff
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is a documentation and coordination tool that does not perform automated execution of commands, network requests, or privilege escalation. It requires explicit user requests for all file operations and uses standard temporary directories for storage.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies a potential attack surface by ingesting external project artifacts into the agent's context. This risk is managed through built-in instructions for the agent:
- Ingestion points: Step 2 of the procedure instructs the agent to read existing artifacts such as specifications, plans, ADRs, and issue trackers.
- Boundary markers: Step 2 requires the agent to clearly separate confirmed decisions from proposals, assumptions, and unresolved questions.
- Capability inventory: The skill utilizes file system read/write operations and repository state inspection via Git.
- Sanitization: Step 3 mandates the redaction of secrets, credentials, and personal information from the generated document and its references.
Audit Metadata