handover
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides instructions for maintaining project documentation, such as Architecture Decision Records (ADRs) and deployment guides, facilitating knowledge transfer.
- [SAFE]: The skill includes proactive instructions for agents to scan their own documentation output for hardcoded secrets before completion, which is a recommended security practice to prevent accidental data exposure.
- [INDIRECT_PROMPT_INJECTION]: The skill processes session data to update documentation files such as CLAUDE.md and AGENTS.md. While this creates a persistence surface for session data, it is considered safe because it is limited to documenting project status and next steps without introducing automated execution of the ingested content.
- Ingestion points: Session history, diffs, and decision candidates (SKILL.md)
- Boundary markers: Absent (SKILL.md)
- Capability inventory: File-system write access to repository documentation and agent configuration files (SKILL.md)
- Sanitization: Absent (SKILL.md)
Audit Metadata