cookbook-migrate-model
Warn
Audited by Gen Agent Trust Hub on Aug 13, 2026
Risk Level: MEDIUMREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: The workflow requires creating and executing a temporary Node.js script to convert legacy JSX logic into configuration data. It also recommends using an equivalence audit script that executes legacy generator code using stubbed React hooks and indirect evaluation.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests legacy documentation and code files (JSX/MDX) to drive code generation.
- Ingestion points: docs/src/snippets/autoregressive/*.jsx and legacy MDX files.
- Boundary markers: None specified.
- Capability inventory: Shell command execution (git, gh, mint), file system writes, and dynamic Node.js script execution.
- Sanitization: None mentioned; the focus is on verbatim transcription of legacy content.
- [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill utilizes shell-based tools including 'git', 'gh', and 'mint' to manage source code and validate documentation. It also suggests checking out specific PRs or commits from GitHub to use as reproducible anchors.
Audit Metadata