All servers / Design / FigJam MCP server
FigJam MCP server
MCP server that turns a FigJam board into queryable context for LLMs
Tested by ConnectATDScore 90 out of 100. Tested 2026-11-19 on package version 0.14.0. Package: figjam-context-mcp · source
Score breakdown
| Runs | 18 / 20 | |
|---|---|---|
| Tool descriptions | 20 / 20 | |
| Package safety | 20 / 20 | |
| Maintenance | 20 / 20 | |
| Adoption | 12 / 20 |
What we saw
- Starts and answers the MCP handshake: yes (2.9 seconds including download)
- Tools offered: 5
- Last published: 2026-11-16
- Weekly downloads (npm): 1,092
- License: MIT · Direct dependencies: 5
Want this server set up for your team?
ConnectATD builds AI automation and wires up MCP servers like this one. Tell us what you are trying to do.
Talk to ConnectATDTools and how they are described
Agents choose tools from these descriptions. Under each name you can see how many of its parameters are explained.
ingest_board7 of 7 parameters describedRead a FigJam/Figma URL with the configured Figma token and persist a local snapshot for get_board_context, answer_from_board and diff_board. The response includes boardId and snapshotId, and its summary names the largest sections when present, counts excerpts outside them and points to get_board_context for the next read. Leaves the Figma file unchanged. Every call downloads the board from Figma (one Tier 1 request); get_board_context and answer_from_board read the saved snapshot without a new request. Balanced/quality modes may send board text and images to the configured LLM provider and incur charges; max_speed skips vision. Re-ingest to refresh stale content and continue pending image interpretation; focusNodeIds with section IDs from the summary or the first page of get_board_context interprets their images first. Local history is bounded. Supports framework phases or customPhases.
get_board_context9 of 9 parameters describedRead source-cited context from an ingested board snapshot. Start overviews, summaries and documentation with a call without topic, nodeIds, limit and maxChars: its first page names the largest sections and gives bounded page/section guidance with IDs; it is not a complete section directory, empty sections may be omitted, and unlisted pages are counted. Its footer gives one reusable read call for the listed page IDs. Then pass section, page, frame or group IDs as nodeIds; several IDs share each page in request order. For multiple pages, read each page separately with detail: "auto", maxChars: 24000 and follow nextCursor with unchanged arguments; page reads include content outside sections, so outsideSections is not required for the same pages. Default detail: "auto" switches reads of multiple distinct IDs or one page, section, frame or group to a compact overview only when the first full page would leave unread excerpts. Overview lines contain node IDs and deterministically shortened text (50–120 characters); structuredContent puts the excerpts once in overview.text, with empty evidence and clusters. limit applies only to the full view and is ignored in an overview, even when explicit. Read the complete source and Figma link with the same boardId and snapshotId, nodeIds: ["<ID>"] and detail: "full"; table IDs read every row. Without overview, excerpts are in evidence. Read outsideSections: true for a separate targeted query, search with topic, or continue with cursor. Cite the node ID or Figma link behind each statement. Synthesize the answer yourself. No network or LLM calls; use ingest_board to refresh Figma content.
answer_from_board3 of 3 parameters describedAnswer one specific question, such as a number, a status or a decision, from the most relevant excerpts of an ingested board, with validated source citations or an insufficient-evidence result. It finds excerpts by the question's words (whole words, ignoring case and accents, plus some word forms and compounds for words of at least five letters; no synonyms or translations), so ask with the board's words and language. Keep its citations (node ID and Figma link) with the statements they support. It reads only a handful of excerpts: not for summaries, lists, risks or next steps across several sections; read those with get_board_context. Sends the question and the selected excerpts to the configured LLM provider; calls may incur charges.
diff_board2 of 2 parameters describedCompare retained local snapshots of the same board, reporting cluster, node and connector changes. Names each changed node with its section and its old and new text, and each changed table cell with its row and column. Reports moves with their origin and destination sections or pages, identifies resizes, and marks added or removed clusters made up of existing nodes as regrouped. Needs two ingests; when the board did not change in between, it reports no changes since the first. compareTo=1 selects the previous snapshot, 2 selects two states back. No network or LLM calls; ingest_board must capture the current Figma state first.
diagnose_llm_configno parametersTest the configured text and vision models with three small synthetic JSON challenges. Sends synthetic text and an image to the configured provider, may incur charges, and reports semantic/schema failures. Does not read board contents. Use after changing model/provider settings.
Tested in a sandbox with an empty environment: no API keys, a throwaway folder, nothing from a real account. A passing test shows the server runs and responds. It is not a security audit.