handoff
Pass
Audited by Gen Agent Trust Hub on Aug 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes conversation history and user requests to generate instructions for a subsequent agent session. This represents a multi-step chain where untrusted data from a previous session can influence the behavior of the next session.
- Ingestion points: Processes the "live conversation" and "user request" as described in the skill's protocol.
- Boundary markers: The generated output skeleton does not include explicit delimiters or instructions to ignore embedded commands in the summarized content.
- Capability inventory: The skill facilitates file referencing (pointing to artifacts like
plan.md) and potential file writing (Step 4 mentions writing files if asked). - Sanitization: No explicit sanitization or filtering of the summarized conversation data is described before it is interpolated into the handoff prompt.
Audit Metadata