claude-handoff
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute the
claudeCLI tool with specific arguments (--bg,--name) to launch background processes. While this is a native platform feature, it involves dynamic command construction based on conversation context. - [INDIRECT_PROMPT_INJECTION]: The skill creates a multi-step chain where the output of the current agent (the handoff summary) is used as the starting prompt for a new background agent.
- Ingestion points: The summary is derived from the current conversation history and user-supplied arguments as defined in
SKILL.md. - Boundary markers: None are present; the summary is passed as a raw string argument to the new agent session.
- Capability inventory: The background agent initiated by this skill possesses full capabilities, including file system access and the ability to execute further commands.
- Sanitization: The skill includes explicit instructions for the agent to redact sensitive information (API keys, passwords, PII) before generating the handoff summary, providing a manual layer of protection.
Audit Metadata