handoff
Pass
Audited by Gen Agent Trust Hub on Jul 7, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acts as a bridge between agents, processing untrusted conversation history into a summary document. This creates a surface for indirect prompt injection if the source conversation contains instructions intended to influence the behavior of the subsequent agent.
- Ingestion points: Current conversation history processed during the summary generation.
- Boundary markers: No specific delimiters or warnings are defined for the output document to help the receiving agent distinguish between the summary and original instructions.
- Capability inventory: File writing capability to the OS temporary directory.
- Sanitization: The skill includes explicit instructions to redact sensitive data such as API keys and passwords.
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill instructions direct the agent to save the summary to the OS temporary directory rather than the local workspace. While this is standard practice for transient data, it involves writing content to locations outside the immediate project environment.
- Evidence: 'Save it to the OS temporary directory — not the current workspace.' (SKILL.md)
Audit Metadata