claude-handoff
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to execute the
claudecommand-line tool to initialize background agents. - [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted conversation history and user arguments to create the initial prompt for a new agent.
- Ingestion points: Conversation history and user-supplied arguments referenced in
SKILL.md. - Boundary markers: The handoff summary is passed as a quoted argument, but no internal delimiters are used to separate user data from instructions.
- Capability inventory: Shell command execution and new process creation via the
claudeCLI. - Sanitization: The agent is explicitly instructed to redact sensitive information such as API keys, passwords, and PII before generating the summary.
Audit Metadata