handoff
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes conversation history (untrusted data) to generate its summary. It incorporates a security control by instructing the agent to redact sensitive information, such as API keys, passwords, and personally identifiable information, before the document is finalized. It also generates suggestions for future skill invocations to assist the next agent.
- [DATA_EXFILTRATION]: The skill writes a file to the operating system's temporary directory outside the current workspace. This is the intended primary purpose of the skill to facilitate session handoffs and does not involve network exfiltration or unauthorized data transmission.
- [COMMAND_EXECUTION]: No shell commands, subprocesses, or external script executions are defined in the instructions or configuration files.
Audit Metadata