handoff
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates a document based on untrusted session history to be read by another agent, which is a potential surface for indirect prompt injection.
- Ingestion points: Processes all content from the current session (SKILL.md).
- Boundary markers: Absent; the skill does not define specific delimiters to separate summarized user content from agent instructions in the output document.
- Capability inventory: The agent requires file system write access to save the handoff document to the OS temporary directory.
- Sanitization: The skill includes an instruction to anonymize sensitive data (API keys, PII), but it does not specify sanitization for executable commands or instructions that might be embedded in the session history.
- [DATA_EXPOSURE_AND_EXFILTRATION]: While the skill processes the entire current session, it proactively addresses data exposure risks by instructing the agent to 'de-identify any sensitive information' such as API keys and passwords before writing to disk.
Audit Metadata