handoffkit
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill inherently processes conversation history, which is untrusted data. While this presents a surface for indirect injection, the skill includes explicit instructions to redact secrets (API keys, tokens, passwords) and PII before outputting the document. It also mandates referencing artifacts by link or path rather than reproducing content, reducing the risk of data leakage.
- [DATA_EXPOSURE]: The skill incorporates a specific 'Redact' procedure to ensure sensitive credentials found in the session context are not persisted into the handoff files. It suggests referring to secrets by location (e.g., '.env file') rather than value.
- [COMMAND_EXECUTION]: The skill uses the 'Write' tool to save documentation to the 'docs/handoffs/' directory. This is a restricted, purposeful use of the filesystem within the workspace scope.
- [SAFE]: The 'disable-model-invocation' configuration is set to true, which limits the agent's autonomy by preventing it from triggering other skills automatically during the handoff process, providing a layer of human-in-the-loop control.
Audit Metadata