The Agent Skills Directory

[SAFE]: The skill uses an external tool to fetch documentation from 'chromatic.com'. This is the official domain of the service described (Chromatic), and the activity is limited to retrieving technical guidelines to assist the user.
[SAFE]: The skill instructions contain explicit boundary constraints that prohibit the agent from committing or pushing files, ensuring that any generated configuration is only presented to the user for review.
[SAFE]: The skill implements a data ingestion pattern by processing external documentation. 1. Ingestion points: URLs listed in 'reference/docs-map.md' pointing to 'chromatic.com'. 2. Boundary markers: The skill instructs the agent to treat fetched content as authoritative but does not define strict delimiters. 3. Capability inventory: The skill is restricted to text generation and forbidden from file writes or commits. 4. Sanitization: No specific sanitization is mentioned. This pattern is evaluated as safe given the trusted nature of the vendor's official documentation source.

chromatic-viewports