measure-mcp

kitchen_measure_vessel

Measure the brimful capacity of a bowl, pot, cake pan, or baking dish from a depth capture or 3D scan to scale recipes accurately.

Instructions

Capacity of a bowl, baking dish, cake pan, sheet pan or pot from one depth capture or 3D scan (to scale a recipe).

Camera looking down into the upright vessel on a flat surface (45 degrees above the table or steeper), whole rim in frame. path: LiDAR/Portrait HEIC, capture bundle (folder or .zip), metric depth file (16-bit PNG in mm, TIFF or NPZ in meters), a plain photo, or a 3D scan: a Scaniverse, Polycam or 3D Scanner App OBJ or PLY (a .zip with the .obj, .mtl and texture is fine). A scan needs the vessel upright on a table with some table in view; it is the object nearest the middle, clutter is ignored, and rgb_path, bbox, point and reference do not apply. Units and up axis are detected and reported in scan and warnings; if wrong, pass scan_units (m|cm|mm|in: the FILE's units; Scaniverse writes meters; units="cm" also works) and up_axis (x|y|z, signs ok). rgb_path: optional original photo (EXIF gives the focal length). bbox [x0, y0, x1, y1] or point [x, y] pick the vessel among several: PHOTO pixels (depth-map pixels when there is no photo). vessel_type: bowl|pot|round_pan|rect_pan|sheet_pan|casserole hint, or auto. material (metal|glass|ceramic|silicone) and wall_thickness_mm set the wall allowance (depth is measured to the outer rim). reference {"type": "credit_card", "corners": [[x,y] x4]}: needed for relative depth, useful when a depth file has no camera intrinsics; see measure_list_reference_objects. units: metric|imperial|both.

A plain photo has no depth. With the base install it returns NO_DEPTH whose hint lists the options in order: LiDAR capture; a credit card in frame plus pip install "measure-mcp[depth]"; typed dimensions (kitchen_vessel_from_dimensions, which always works). measure_depth_info shows what is installed.

Returns summary (quote it), vessel (pass it as source_vessel/target_vessel to kitchen_scale_recipe), measurements (rim size, inner depth, brimful capacity, each +/- about 95% with a confidence), capacity (L, mL, cups, quarts), fill_guidance, nearest_preset, quality (fit_grade, fit_issues) and warnings. Capacity is BRIMFUL, inner surface. Confidence is never "high" for relative or monocular depth, a noisy or poorly fitting capture, or guessed camera intrinsics (those are "low"); a rim that fits no clean circle/rectangle returns VESSEL_FIT_POOR instead of a number. Present it as an estimate; typed dimensions are more exact.

Input Schema

TableJSON Schema

Name	Required	Default
`bbox`	No
`path`	Yes
`point`	No
`units`	No	both
`up_axis`	No	auto
`material`	No	unknown
`rgb_path`	No
`reference`	No
`scan_units`	No	auto
`vessel_type`	No	auto
`depth_source`	No	auto
`wall_thickness_mm`	No

Output Schema

TableJSON Schema

Name	Required	Description	Default
No arguments

Schema Changelog

Changes observed during successful MCP inspections.

First observedv0.1.0Oct 16, 2026

TDQS

A4.8/5.0

Behavior5/5

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

No annotations are provided, so the description carries the full burden and does so unusually well: it discloses supported input families, that units/up-axis are auto-detected and where the result is reported, that capacity is BRIMFUL measured to the inner surface, the confidence grading rules (never 'high' for relative/monocular/guessed intrinsics), and explicit error codes (NO_DEPTH, VESSEL_FIT_POOR).

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?

Every fact is relevant, but the description is a dense wall of run-on sentences with embedded parentheticals and backticks that is hard to scan. Several sentences enumerate return fields (summary, vessel, measurements, capacity, fill_guidance, nearest_preset, quality, warnings) that an output schema already conveys, so some length is not earning its place.

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 12-parameter tool spanning several capture formats, this covers everything an agent needs to attempt a correct call: required capture geometry, format support, disambiguation of the vessel among several, unit/axis overrides, fallback behavior, and how to interpret the estimate. The output schema handles return structure, so the description's completeness is more than sufficient.

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 description coverage is 0%, so the description must compensate and largely does: path formats, rgb_path (EXIF focal length), bbox/point pixel coordinate spaces (photo vs depth-map), vessel_type and material enums, wall_thickness_mm's effect on the allowance, reference's credit-card corner structure with a cross-reference to measure_list_reference_objects, scan_units (file units, Scaniverse=meters) and up_axis signs. Only depth_source goes unexplained.

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?

Opens with a specific verb-and-resource statement: capacity of a named set of vessels (bowl, baking dish, cake pan, sheet pan, pot) from one depth capture or 3D scan, with the motivation (scaling a recipe). It is clearly distinguishable from kitchen_vessel_from_dimensions (typed dimensions) and kitchen_scale_recipe (downstream consumer), which it names.

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 concrete capture preconditions (camera 45 degrees or steeper, whole rim in frame, vessel upright on a table) and explicitly routes around the failure path: a plain photo returns NO_DEPTH with an ordered hint list, and typed dimensions via kitchen_vessel_from_dimensions 'always work'. It even names measure_depth_info as the way to check installed capabilities.

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