claude-handoff
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the
claudeCLI to spawn background processes (claude --bg). It utilizes a secure argument-passing pattern ("$(cat <summary file>)") specifically designed to prevent the shell from interpreting or executing malicious characters (like backticks or variable expansions) that might be present in the conversation summary. - [INDIRECT_PROMPT_INJECTION]: The skill generates a summary from the current conversation history, which could theoretically contain untrusted data.
- Ingestion points: Conversation history (referenced in SKILL.md).
- Boundary markers: The instructions explicitly command the agent to redact sensitive information and to reference rather than duplicate existing technical artifacts.
- Capability inventory: Spawning background agent processes via the local
claudeCLI. - Sanitization: The skill mandates the redaction of sensitive information, such as API keys, passwords, and personally identifiable information (PII), before the summary is used as a prompt for the next agent.
Audit Metadata