Skip to main content
Glama
the-real-talk

the-real-talk

audit_bindings

Audit Figma node bindings under a root node against variable, spacing, padding, and height rules to return unbound or noncompliant design values.

Instructions

Audit the bindings under rootId against the measure rules, computed inside the plugin. DEFAULT RULES (all five apply when rules is omitted): (1) unbound: a spacing, padding, fixed width or height, radius, stroke weight, SOLID fill/stroke color or text without style that has no variable bound (itemSpacing counts only with 2 or more visible in-flow children and not with SPACE_BETWEEN; padding 0 and radius 0 are not flagged here; width and height count only when the sizing is FIXED); (2) primitive: a value bound to a variable of the primitive collection instead of a role variable; (3) not_multiple_of_4: itemSpacing, counterAxisSpacing, padding, fixed width and height, or a line height in px that is not a multiple of 4 (strokes and radii excluded; a width or height from 1 to 3 is treated as a hairline and skipped); (4) zero_padding: an auto layout frame, component or instance with children and a visible fill or stroke whose padding is 0 (one finding per side); (5) height_not_in_set: a height outside the default set 24, 32, 36, 44, 56, 64 (tolerance 0.0005; text, vectors, groups and similar nodes are skipped). Only the findings come back, as rows [nodeId, name, field, value, variable, rule] (see columns), after a totals-per-rule header. rules (default all five): unbound (spacing, padding, fixed width and height, radius, stroke weight, SOLID fill/stroke color or text without style, with no variable: itemSpacing counts when 2 or more children, padding 0 and radius 0 are not flagged here), primitive (bound to a variable of the collection primitiveCollection, default "Primitives", instead of a role), not_multiple_of_4 (itemSpacing, counterAxisSpacing, padding, fixed width and height, line height in px; strokes and radii excluded), zero_padding (auto layout frame with children and a visible fill or stroke whose padding is 0; one finding per side), height_not_in_set (height outside heightSet, default 24, 32, 36, 44, 56, 64; heightScope "components" (default) checks COMPONENT nodes and nodes whose height is bound, "all" checks every node but text, vectors, groups). Instance children are excluded unless excludeInstanceChildren is false. maxResults (default 200, max 5000; 0 = counts only, scans as far as the time budget goes) stops the scan when reached: PARTIAL with nextCursor, totals cover the nodes visited by this call only, add them up across calls. includeZeroRadius:true also flags an unbound radius of 0. Read only.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "rootId": "<rootId>"
    },
    "name": "audit_bindings"
  }
}

Input Schema

NameRequiredDescriptionDefault
rulesNo
cursorNonextCursor of a previous PARTIAL answer, to continue where it stopped
rootIdYes
heightSetNo
maxResultsNo
heightScopeNo
timeBudgetMsNoTime budget of the plugin for this call in ms (default 60000, max 100000). When it is reached the answer says it is PARTIAL and gives nextCursor.
includeZeroRadiusNo
primitiveCollectionNo
excludeInstanceChildrenNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv1.0.0

TDQS

A4.4/5.0
Behavior5/5

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

With no annotations at all, the description carries the full burden and does so well: it declares the tool read-only, states computation happens inside the plugin, discloses that maxResults truncation yields PARTIAL with nextCursor and that totals only cover visited nodes (must be summed across calls), gives the 0.0005 height tolerance, names skipped node types, and specifies the row/header return shape.

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 purpose is front-loaded and the parenthetical rule clauses are scannable, but the five rules are explained twice — once under 'DEFAULT RULES' and again under the 'rules' parameter — with substantial near-verbatim repetition. Roughly half the text is redundant, which is a real cost in a single paragraph of this length.

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 10-parameter, annotation-free tool with no output schema, the description covers what is needed: selection rules, defaults, scoping, truncation/continuation behavior, and the returned row and header format. Nothing an agent needs to call it correctly or interpret the response appears to be missing.

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

Parameters5/5

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

Schema coverage is only 20% and the description compensates for nearly every parameter: defaults and meaning for rules, heightSet (24/32/36/44/56/64), heightScope ('components' vs 'all'), primitiveCollection ('Primitives'), maxResults (200, max 5000, 0 = counts only), includeZeroRadius, excludeInstanceChildren and cursor continuation. This adds real behavioral meaning well beyond the bare schema types.

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 ('Audit the bindings under rootId against the measure rules, computed inside the plugin') and then enumerates all five rules, so the agent knows this is a rule-based linter rather than a plain binding getter like get_nodes_bindings. The purpose is unmistakable without opening the schema.

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

Usage Guidelines3/5

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

Usage is implied rather than stated: the agent learns that omitting rules applies all five, that rules selects a subset, and that cursor continues a PARTIAL scan. There is no explicit 'use this instead of X' routing to the sibling binding tools and no when-not guidance, leaving the agent to infer the boundary.

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

Deploy Server

Other Tools