Tools provided
Tools the agent exposes — your AI client calls them automatically when it needs them.
- run_full — Orchestrated path: match the provided Figma node against your design tokens and component library, then generate the Vue or React component, the Storybook story and the snapshot test as a PR draft.
- analyze — Run pure deterministic analysis over the provided Figma node, tokens and existing components with no model call, returning scored token, reuse and accessibility findings.
- match_tokens — Match every raw value on every layer against your design tokens by exact value equality — the token-mapping step on its own. No fuzzy matching.
- list_hardcoded_values — Return only the raw values with no exact token match — the hardcoded-risk list, emitted as raw values rather than tokenized to something merely close.
- detect_reused_components — Detect which of your existing library components this node's layers can reuse.
- list_findings — Return only the scored findings — token coverage, component reuse and accessibility gaps — for the provided node.
- list_capabilities — List the agent's static capabilities: the available tools and the supported frameworks. Useful for discovery.
- plan_inputs — Plan the inputs for a tool: returns the clarifying questions, the JSON schema and a ready-to-edit example.
- discover_intent — Understand your goal and co-design the exact input through clarifying questions before running.
- open_form — Opens the guided form to collect the Figma link, the target path and the framework. The entry point for running the agent on your own node.
- run_form — Internal: invoked by the guided form when the user submits. Not for direct use.
