handoff-prompt
Pass
Audited by Gen Agent Trust Hub on Apr 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a documentation and templating guide for prompt engineering. It does not include any code, shell commands, or network requests.\n- [DATA_EXFILTRATION]: The skill guides the agent to include source code and architecture details in a prompt for delegation. This is the intended purpose of the skill and relies on the user to manually move the generated text, presenting no automated exfiltration risk.\n- [PROMPT_INJECTION]: The process of ingesting codebase files to generate a handoff prompt creates a surface for indirect prompt injection if those files contain malicious instructions.\n
- Ingestion points: Local codebase files read during the drafting process (e.g., source code, file inventory).\n
- Boundary markers: The skill recommends using Markdown code blocks to delimit code snippets.\n
- Capability inventory: The skill has no capabilities to execute commands or interact with the filesystem/network beyond text generation.\n
- Sanitization: No explicit sanitization or filtering of codebase content is mentioned.
Audit Metadata