promptbook-migrate
Pass
Audited by Gen Agent Trust Hub on Jul 4, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because its primary function is to ingest and process user-provided source code and prompt files (as described in the 'Inventory' and 'Pick the source playbook' sections). Malicious instructions embedded in the processed prompts could potentially influence the agent's behavior during migration.\n
- Ingestion points: The skill instructions guide the agent to read existing project files (TypeScript, JavaScript, Markdown, and Text) using search tools like
rgto find and extract prompt literals.\n - Boundary markers: There are no explicit instructions to use delimiters or ignore embedded instructions when reading user-provided content.\n
- Capability inventory: The skill requires file-read, file-write, and code modification capabilities to create fragments, rule files, and replace code strings with
resolve()calls.\n - Sanitization: No explicit sanitization or validation of the ingested prompt content is defined in the instructions.\n- [EXTERNAL_DOWNLOADS]: The skill mentions and utilizes the '@markbrutx/promptbook-core' library and the 'promptbook' CLI tool. These are recognized as legitimate resources provided by the skill's author ('markbrutx').
Audit Metadata