Skip to main content
Glama

Review Asset

lab_review_asset
Idempotent

Review generated assets: set verdict (accepted|rejected|pending), selected and/or defect tags. Pass ONE id, or ids for a bulk review — after a generation run there are usually dozens to sort, and clicking through them one by one is why the review fields went unused. Rejected assets drop out of the default listings and become prunable. Note: single-id calls make selected exclusive per scene (one chosen variant per scene_ref); bulk calls do NOT, because a bulk over one scene would keep resetting itself and only the last id would stay selected. To record WHAT is wrong, pass tags (see the field) — with or without a verdict; a call with only tags changes no verdict. Tags are written after the verdict: if the tags fail, the verdict stays saved and the answer says so under tags_error. Bulk tagging is not available on a device grant.

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": {
      "reference_type": "<reference_type>"
    },
    "name": "lab_review_asset"
  }
}

Input Schema

NameRequiredDescriptionDefault
idNoSingle asset id (positive integer as string). Use this for the per-scene selection
idsNoMany asset ids for a bulk review (max 200). Mutually exclusive with `id`
tagsNoDefect tags: WHAT is wrong, from a fixed list — independent of the verdict (an accepted asset can carry tags; tags never change verdict or selected). At most 5 per asset. Pass only tags that apply to the asset type (the brackets show where each applies). action_wrong [video/image] — a requested action does not happen or happens differently (also ambient movement such as shadows, lights, crowds); expression_wrong [video/image] — facial expression, gaze or blinking is wrong; camera_motion [video] — the camera moves differently from the request (handheld shake or push-in missing, wrong direction); framing [video/image] — shot size, angle, horizon or position in the frame differs from the request; colour_shift [video/image] — colour or light differs from the request or jumps; wrong_count [video/image] — wrong number of people or things (e.g. fewer extras than asked); missing_element [video/image] — a requested object or detail is not there; identity_drift [video/image] — face, wardrobe or build differs from the reference; wrong_place [video/image] — the place or layout does not match; anachronism [video/image] — something from the wrong era in a period setting; extra_people [video/image] — a person who is not part of the cast; text_wrong [video/image] — requested text is wrong or unreadable; anatomy [video/image] — hands, limbs or face are distorted; artifacts [video/image] — flicker, distortion, watermark, logo or a visible seam; frozen [video] — the picture freezes; continuity_jump [video] — the join at a cut or chained clip jumps; lip_sync [video] — lips do not match the sound; audio_wrong [video/sound] — sound is missing, jumps or does not fit; other [video/image/sound] — anything else — put the detail in `reason`.
reasonNoWhy — stored as verdict_reason, useful when reviewing a rejection later
verdictNoAcceptance verdict. Feeds the acceptance telemetry (which provider/mode/price actually produced usable shots)
selectedNoMark as the chosen variant
tags_modeNoadd (default): add to the tags this person already set on the asset. replace: set exactly these tags (an empty list clears them). Only the caller's own tags are touched, never another person's or the machine tags.
reference_typeYesWhat kind of asset these ids refer to

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed2 schema fields changed
    • addedInput schema / properties / tags
      Added value: +{
      +  "description": "Defect tags: WHAT is wrong, from a fixed list — independent of the verdict (an accepted asset can carry tags; tags never change verdict or selected). At most 5 per asset. Pass only tags that apply to the asset type (the brackets show where each applies). action_wrong [video/image] — a requested action does not happen or happens differently (also ambient movement such as shadows, lights, crowds); expression_wrong [video/image] — facial expression, gaze or blinking is wrong; camera_motion [video] — the camera moves differently from the request (handheld shake or push-in missing, wrong direction); framing [video/image] — shot size, angle, horizon or position in the frame differs from the request; colour_shift [video/image] — colour or light differs from the request or jumps; wrong_count [video/image] — wrong number of people or things (e.g. fewer extras than asked); missing_element [video/image] — a requested object or detail is not there; identity_drift [video/image] — face, wardrobe or build differs from the reference; wrong_place [video/image] — the place or layout does not match; anachronism [video/image] — something from the wrong era in a period setting; extra_people [video/image] — a person who is not part of the cast; text_wrong [video/image] — requested text is wrong or unreadable; anatomy [video/image] — hands, limbs or face are distorted; artifacts [video/image] — flicker, distortion, watermark, logo or a visible seam; frozen [video] — the picture freezes; continuity_jump [video] — the join at a cut or chained clip jumps; lip_sync [video] — lips do not match the sound; audio_wrong [video/sound] — sound is missing, jumps or does not fit; other [video/image/sound] — anything else — put the detail in `reason`.",
      +  "items": {
      +    "enum": [
      +      "action_wrong",
      +      "expression_wrong",
      +      "camera_motion",
      +      "framing",
      +      "colour_shift",
      +      "wrong_count",
      +      "missing_element",
      +      "identity_drift",
      +      "wrong_place",
      +      "anachronism",
      +      "extra_people",
      +      "text_wrong",
      +      "anatomy",
      +      "artifacts",
      +      "frozen",
      +      "continuity_jump",
      +      "lip_sync",
      +      "audio_wrong",
      +      "other"
      +    ],
      +    "type": "string"
      +  },
      +  "maxItems": 5,
      +  "type": "array"
      +}
    • addedInput schema / properties / tags_mode
      Added value: +{
      +  "description": "add (default): add to the tags this person already set on the asset. replace: set exactly these tags (an empty list clears them). Only the caller's own tags are touched, never another person's or the machine tags.",
      +  "enum": [
      +    "add",
      +    "replace"
      +  ],
      +  "type": "string"
      +}
  2. Changed1 schema field changed
    • addedInput schema / additionalProperties
      Added value: +false
  3. First observed

TDQS

A4.5/5.0
Behavior5/5

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

The description adds substantial behavior beyond the annotations: rejected assets drop out of default listings and become prunable, single-id selection is exclusive per scene while bulk selection is not, tags are written after the verdict with partial-failure reporting under `tags_error`, and bulk tagging is unavailable on a device grant. These are exactly the operational details an agent needs before invoking a mutation.

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

Conciseness4/5

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

The description is front-loaded with the core action and bulk/single distinction, and most sentences carry operational information. It is somewhat long and includes a motivational aside about clicking through assets one by one, but the detail is mostly justified by the complexity of the tool.

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

Completeness4/5

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

For a complex mutation tool with no output schema, the description covers the important behavioral caveats: selection semantics, bulk limitations, tag write ordering, and partial failure. It does not fully describe the response shape beyond `tags_error`, but overall it gives enough context for correct invocation.

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 the schema already documents all eight parameters thoroughly. The description still adds valuable cross-parameter semantics, especially the single-id `selected` exclusivity versus bulk behavior and the rule that tags alone change no verdict. It does not add much beyond the schema for parameters like `reason` or `reference_type`, so a 4 is appropriate.

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?

The description states a specific verb and resource: review generated assets and set verdict, selected state, and defect tags. It immediately distinguishes single-id review from bulk review via `ids`, so an agent can understand the tool's scope 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 Guidelines4/5

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

It gives clear usage context: pass one `id` for per-scene selection, or `ids` for bulk review after a generation run where dozens of assets need sorting. It also explains that tags can be passed with or without a verdict. It does not explicitly name sibling alternatives such as lab_review_queue, so it falls short of a 5.

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources