Skip to main content
Glama

figma_inspect

Read-onlyIdempotent

Reads live Figma nodes in one call, returning compact outlines, exact properties, copy locations, and diffs against a reference node so you can inspect and verify designs without changing the file.

Instructions

READ the live Figma document in ONE call — the tool for any read (not figma_execute_code); pass every node id you need at once instead of one call per node. Default: a compact outline, one line per node (TYPE "name" #id WxH @x,y [V gap8 pad16 fill/hug a:start/center] size:fill/hug fill:#FFF stroke:#E8EAF0/0,1,1,1 r0,0,3,3 shadow:… font "text…"; non-defaults only; a TEXT named after its text drops the name; a:P/C = primaryAxisAlignItems/counterAxisAlignItems, start = MIN, center = CENTER, end = MAX, between = SPACE_BETWEEN, absent = MIN/MIN), as deep as fits the reply unless depth is given; children:N marks a node with more below. like: "<ref id>" (or several) FINDS every copy of the reference under node_ids (screens/sections; default: its section) — instances of the same component, or same name — and diffs each against it: the call for "redo these like the reference" when you don't know where the copies are. compare: "<ref id>" lists, per node_id, every difference from that reference (layout, sizing, stroke, radius, effects, clip, font, variant; children matched by name) and ends with REF lines, the reference's exact values to copy into the write (no props read needed) — one call for "make these like that one", and the check after the write. props returns exact values instead: { id: { prop: value } } (also fill, stroke, text, font, layout, padding, parent, children, main, variant, props, absolute, reactions, connector). find searches node names under node_ids (or the current page); find_text searches text content. view: "map" = canvas map (sections, breakpoints). view: "flow" = the prototype links under node_ids (a section or screens): per screen, source, trigger, action, target, transition, plus the overlay settings of overlay targets — never scan the page for reactions. context: true adds ancestor sections, breakpoints and component variants. offset skips the first N outline lines when a reply ends with pass offset=K. check returns only mismatches. Combine them freely. Changes nothing.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "figma_inspect"
  }
}

Input Schema

NameRequiredDescriptionDefault
idsNoAlias of node_ids.
findNoCase-insensitive substring of node names to search for under node_ids (or the current page). Up to 8 matches come back with their subtree outline (as deep as fits, no extra read needed); more matches as one line each.
likeNoReference node id (several: comma-separated): find its copies under node_ids (default: the reference's top section) and list, per copy, its frame and how it differs (missing children come with the reference's own line, and REF lines at the end give the reference's exact values, so no extra read is needed). Copies that already match are listed as identical.
viewNo"outline" (default) = one line per node; "map" = canvas map: sections, frames, breakpoints and component sets in a few lines; "table" = per table (found by structure) the columns: header, width, sizing, cell contents, plus row and cell ids; "flow" = prototype links (screen, source, trigger, action, target, transition, overlay settings) under node_ids (without node_ids: every top-level screen of the current page).
checkNoExpectations to verify: { "12:34": { "width": 320, "fill": "#FFFFFF", "layout": "V gap8 pad16" } }. Returns { pass, fail: [{ id, key, want, got }], missing }.
depthNoOutline levels below each node (0 = the nodes only, max 6). Omit it: the outline goes as deep as fits the reply.
limitNoMax find matches (default 30).
propsNoReturn these properties per node instead of an outline, e.g. ["width", "layoutSizingHorizontal", "fill", "text"]; also "reactions" (prototype interactions), "connector" (FigJam connector ends) and "texts" (all texts below a frame, id + full characters, pages with offset).
offsetNoSkip the first N lines of the outline / found list (use the K from a '… pass offset=K' line to read the next page).
targetNofileName of a specific connected Figma document, when more than one is open (see figma_list_targets).
compareNoReference node id: instead of an outline, list how each node_id differs from it (x/y/width and text content are skipped; height, stroke sides/align are compared). Ends with REF lines: the exact values of the differing reference nodes, to copy in figma_execute_code without a props read. 'identical' = nothing to fix.
contextNoAlso return, per node: ancestor sections, breakpoint siblings and component variants ({ id: context }).
node_idNoAlias of node_ids for one id.
node_idsNoNode ids to read ("12:34" or "12-34"). Omit to use the current page.
find_textNoCase-insensitive substring of TEXT content to search for under node_ids (or the current page). One line per match: #id "layer" «text» in <parent layer #id>, screen "<top frame>" (the parent is the frame to click or edit).
find_typeNoRestrict find to one node type, e.g. FRAME, TEXT, INSTANCE.
max_output_bytesNoCap on the text response in UTF-8 bytes (default 3500; in clients that spill big outputs to a file, 0 and values above 3900 become 3900). A cut outline ends with '… pass offset=K': use `offset` to page instead of raising this.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changedv4.4.0
    • changedInput schema / properties / props / description
      Previous value: -"Return these properties per node instead of an outline, e.g. [\"width\", \"layoutSizingHorizontal\", \"fill\", \"text\"]; also \"reactions\" (prototype interactions) and \"connector\" (FigJam connector ends)."New value: +"Return these properties per node instead of an outline, e.g. [\"width\", \"layoutSizingHorizontal\", \"fill\", \"text\"]; also \"reactions\" (prototype interactions), \"connector\" (FigJam connector ends) and \"texts\" (all texts below a frame, id + full characters, pages with offset)."
  2. First observedv0.1.0

TDQS

A4.5/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint/destructiveHint=false/idempotent, so the safety profile is covered; the description still adds value by stating 'Changes nothing', disclosing truncation and paging behavior (max_output_bytes, 'pass offset=K'), the default depth-fits-the-reply rule, and the `children:N` truncation marker. It stops short of describing cost or limits per mode.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The critical routing sentence is front-loaded, but everything after it is one very long run-on block mixing outline grammar, mode semantics and paging rules with backticks and parentheses. The information is high-value yet hard to scan; the outline grammar in particular would parse better split from the mode selection.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 17-parameter, multi-mode tool with no output schema, the description supplies the return shape of every mode (outline line grammar, compare REF lines, flow link fields, props shape { id: { prop: value } }) plus paging behavior. An agent has enough to call it and interpret the reply without a trial call.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so baseline is 3; the description goes beyond it by framing modes in agent-intent terms ('the call for "redo these like the reference" when you don't know where the copies are') and explaining how `props`, `find`, `find_text`, `context` and `check` interact and combine. The dense outline-grammar explanation covers return format more than parameters, so it does not lift this to a 5.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('READ the live Figma document in ONE call') and immediately distinguishes itself from the sibling it is not ('not figma_execute_code'). An agent can tell it apart from figma_read_canvas or figma_screenshot by the explicit read-anything scope.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives explicit routing: pass all node ids at once, use `like` for 'redo these like the reference', `compare` for 'make these like that one' and as the post-write check, `view: flow` 'never scan the page for reactions', `offset` when a reply ends with 'pass offset=K'. Alternatives and when-not conditions are named rather than implied.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.