The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a bundled helper script vrt.mjs using the node interpreter to perform core VRT logic, including image processing and Playwright-based browser automation.
[EXTERNAL_DOWNLOADS]: The skill makes network requests to the Figma API (api.figma.com) to fetch design images and asset metadata. It also uses playwright, which typically manages and downloads browser binaries.
[PROMPT_INJECTION]: The skill's workflow creates an indirect prompt injection surface because it ingests and processes content from external sources that can be influenced by an attacker.
Ingestion points: Data retrieved from the Figma API and the visual output of the rendered Storybook story.
Boundary markers: Absent; the instructions do not define specific delimiters or instructions to ignore embedded text or patterns within the fetched design or code rendering.
Capability inventory: The agent has the capability to read/write files and execute the Node.js helper script across the local filesystem.
Sanitization: There is no sanitization or validation of the text or visual patterns retrieved from Figma before the agent performs a "qualitative read" of the diff image.
[EXTERNAL_DOWNLOADS]: The instructions recommend including an external font stylesheet from rsms.me in the Storybook preview configuration to ensure consistent font rendering between Figma and the implementation.

fe-design-verify