The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches version manifests from 'styleseed-demo.vercel.app' and update archives from the author's GitHub repository. These sources are recognized as legitimate vendor resources hosted on well-known services.
[COMMAND_EXECUTION]: Executes diagnostic scripts and npx commands to perform updates and verify installation integrity. These operations are limited to the skill's management scope.
[INDIRECT_PROMPT_INJECTION]: The skill processes remote JSON metadata that influences update command parameters, representing a potential vulnerability surface.
1. Ingestion points: Fetches remote JSON manifests in 'scripts/check-update.mjs'.
1. Boundary markers: The skill instructions explicitly caution against automatic execution of remote metadata and emphasize manual revision verification.
1. Capability inventory: Accesses Read, Write, Edit, and Bash tools.
1. Sanitization: Uses structured JSON parsing and performs distribution integrity checks via the 'verifyDistribution' function.

ss-update