MCP server for transforming Figma designs into production-ready code. Converts Figma nodes to HTML, React components, and generates UI code with design-to-code workflow integration.
F2C MCP has 5 tools with Figma design-to-code focus. Tool naming follows verb_noun pattern (figma_to_code, figma_get_file_data, etc.), which is good. However, descriptions vary widely in quality: some are detailed and contextual (figma_to_code: 192 chars), others are minimal (figma_get_file_data: 38 chars). All tools have input schemas with types and parameter descriptions, but schema documentation is uneven. Notable gaps: (1) no documented output schemas for any tool, (2) missing error handling guidance, (3) no mention of pagination or result limits despite tools potentially returning large datasets, (4) parameter descriptions lack actionable constraints (e.g., format requirements, valid ranges).
Get detailed information about a Figma file
Get images of Figma nodes
Transform Figma designs into production-ready code. This tool converts selected Figma nodes into HTML,enabling seamless design-to-code workflow.
Generate UI code for a given node or the currently selected node in the Figma desktop app. Use the nodeId parameter to specify a node id. If no node id is provided, the currently selected node will be used. If a URL is provided, extract the node id from the URL, for example, if given the URL https://figma.com/design/:fileKey/:fileName?node-id=1-2, the extracted nodeId would be 1:2.
Generate an image for a given node or the currently selected node in the Figma desktop app. Use the nodeId parameter to specify a node id. If no node id is provided, the currently selected node will be used. If a URL is provided, extract the node id from the URL, for example, if given the URL https://figma.com/design/:fileKey/:fileName?node-id=1-2, the extracted nodeId would be 1:2. Also export Figma design images for visual verification and design fidelity validation. Essential for comparing generated code output against original designs, ensuring pixel-perfect implementation and catching visual discrepancies during the design-to-code process.
figma_get_file_data has minimal description (38 chars: 'Get detailed information about a Figma file'). Below 50-char baseline. LLM cannot distinguish when to call this vs figma_get_images or figma_to_code without reading parameter names.
No documented output schemas for any of the 5 tools. LLM cannot plan downstream operations or extract required fields. Pattern baseline: 100% of A+ tools have documented return types.
Parameter descriptions lack actionable constraints. Examples: 'ids' parameter described as 'Comma-separated list' but no max count or length limit stated. 'personalToken' optional field lacks guidance on precedence vs server-side token. 'depth' parameter in figma_get_file_data has no range (min/max) specified.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | C | 61 | <=2025-11-25 | v2 |
| 2026-03-09 | D | 51 | - | v1 |
No error handling guidance in tool descriptions. If a Figma file is not found, or personalToken is invalid, tools do not document recovery steps. Pattern requires: 'Error responses must tell the LLM what to do next.' None of the 5 tools include this.
Duplicate tool pairs with overlapping responsibilities: figma_to_code vs get_code (both convert Figma nodes to code), figma_get_images vs get_image (both export images). LLM must reason which to call. Pattern: 'Avoid multiple tools that do the same thing differently.'
personalToken parameter repeated across all 5 tools with identical description. Pattern: secret-injection forbids credentials as tool params. Should use server-side injection via figmaConfig (as code shows) and not expose to LLM. Current approach risks token leakage in logs.