handoff
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the current conversation history to generate a handoff document designed to influence the behavior of a future agent session.
- Ingestion points: The skill processes the entire "current conversation" as its input source (SKILL.md).
- Boundary markers: No explicit delimiters or instructions are provided to help the agent distinguish between legitimate conversation data and potentially malicious instructions embedded within that data.
- Capability inventory: The skill is capable of writing the generated summary to the user's OS temporary directory.
- Sanitization: Although the instructions explicitly call for the redaction of sensitive information (API keys, passwords, PII), there are no measures to sanitize the content for prompt injection patterns that might override the next agent's behavior.
Audit Metadata