handoff
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill explicitly requires the agent to redact sensitive data such as API keys and passwords before saving the handoff document.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted conversation history. Ingestion points: Conversation context in SKILL.md. Boundary markers: Not specified. Capability inventory: Writing a summary file to a temporary directory. Sanitization: Explicit redaction of sensitive info. The risk is mitigated by the 'disable-model-invocation' setting which prevents the agent from calling tools during execution.
- [SAFE]: The policy configuration in agents/openai.yaml prevents implicit invocation, ensuring the skill is only run with user consent.
Audit Metadata