migrate
Pass
Audited by Gen Agent Trust Hub on May 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill performs local site transformation and asset management within a dedicated project directory. Analysis confirms it does not perform unauthorized network communication, handle sensitive credentials, or employ obfuscation techniques.
- [COMMAND_EXECUTION]: The skill integrates with command-line tools like '$stardust' and '$impeccable' to manage the migration lifecycle and audit generated output. These integrated tools are used for project initialization and validation.
- [PROMPT_INJECTION]: The skill uses an LLM to process site content and archetypes, creating an indirect prompt injection surface. Ingestion points: Website data is ingested from 'current/pages/.json' and archetype structures from 'stardust/prototypes/'. Boundary markers: Processing is guided by validation contracts and structural provenance metadata. Capability inventory: The skill writes HTML and sidecar JSON metadata to the 'stardust/migrated/' directory and updates the local state file. Sanitization: The rendering process removes third-party scripts, analytics tags, and tracking pixels from the final output.
Audit Metadata