LeanrIQ

Design, render and save a review-held editable scratch Display working version using Creative OS.

build_creative_os

Idempotent

Design, render and save a review-held editable scratch Display working version using Creative OS.. Classification: writes to Studio, creates billable AI usage. Execution: durable. Task support: optional. Mutation modes: validate_only, preview_only, execute. Partial failure policy: completed_items_not_rolled_back. The returned JSON is internal application state for LeanrIQ and ChatGPT orchestration. Use assistantSummary first when responding to the user. Never display raw JSON, data, or meta unless the user explicitly asks for raw JSON. Mention IDs only when useful for follow-up or debugging. Prefer concise, human language, highlight the most useful insight, and suggest the next best LeanrIQ action. If idempotencyKey is omitted on a mutation, the MCP layer derives a deterministic retry key from the tool and business arguments so an accidental identical replay cannot create duplicate work. For a mutation when more than one seat is authorised, seatId or seatName must be supplied on that exact tool call; a remembered or previously selected seat is not sufficient for a write.

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": {
      "accountId": "<accountId>",
      "advertiserId": "<advertiserId>",
      "campaignId": "<campaignId>",
      "creativeId": "<creativeId>",
      "idempotencyKey": "<idempotencyKey>"
    },
    "name": "build_creative_os"
  }
}

Input Schema

Name	Required	Default
`engine`	No
`seatId`	No
`seatName`	No
`accountId`	Yes
`campaignId`	Yes
`creativeId`	Yes
`qualityMode`	No	STANDARD
`advertiserId`	Yes
`sourceLocale`	No
`targetMarket`	No
`contentLocale`	No
`marketContext`	No
`targetMarkets`	No
`contentLocales`	No
`idempotencyKey`	Yes
`premiumBudgetUsd`	No
`blankEditorSource`	No
`imageEditorSource`	No
`reassessVersionId`	No
`defaultContentLocale`	No
`marketLanguageProfile`	No
`resolvedContentLocale`	No
`languageVariantGroupId`	No
`requestedOutputLocales`	No
`supportedContentLocales`	No

Output Schema

Name	Required	Description	Default
`data`	Yes
`meta`	Yes
`error`	No
`status`	Yes
`assistantSummary`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

Changed2 schema fields changedNov 8, 2026

addedInput schema / properties / blankEditorSource

Added value: +{
+  "additionalProperties": false,
+  "properties": {
+    "documentDigest": {
+      "pattern": "^[a-f0-9]{64}$",
+      "type": "string"
+    },
+    "versionId": {
+      "maxLength": 240,
+      "minLength": 1,
+      "type": "string"
+    }
+  },
+  "required": [
+    "versionId",
+    "documentDigest"
+  ],
+  "type": "object"
+}

addedInput schema / properties / imageEditorSource

Added value: +{
+  "additionalProperties": false,
+  "properties": {
+    "documentDigest": {
+      "pattern": "^[a-f0-9]{64}$",
+      "type": "string"
+    },
+    "versionId": {
+      "maxLength": 240,
+      "minLength": 1,
+      "type": "string"
+    }
+  },
+  "required": [
+    "versionId",
+    "documentDigest"
+  ],
+  "type": "object"
+}

Changed3 schema fields changedNov 7, 2026

removedInput schema / properties / engine / default
```
Removed value: -"original"
```

addedInput schema / properties / premiumBudgetUsd

Added value: +{
+  "exclusiveMinimum": 0,
+  "maximum": 100,
+  "type": "number"
+}

addedInput schema / properties / qualityMode

Added value: +{
+  "default": "STANDARD",
+  "enum": [
+    "STANDARD",
+    "PREMIUM_CREATIVE"
+  ],
+  "type": "string"
+}

AddedNov 6, 2026

TDQS

B3/5.0

Behavior4/5

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

Annotations only declare readOnly=false, idempotent=true, destructive=false, openWorld=false. The description adds genuinely non-obvious behavior: it writes to Studio, consumes billable AI budget, execution is durable, partial failures are NOT rolled back (completed_items_not_rolled_back), and there are three mutation modes. The idempotency-key derivation rule also clarifies retry safety beyond the idempotentHint. It stops short of saying whether the scratch version is later promoted or discarded, which matters for a "review-held" artifact.

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 sentence is front-loaded, which is good, but the body allocates a large share of text to response-formatting boilerplate ("Use assistantSummary first", "Never display raw JSON") rather than tool behavior, and there is a doubled period plus an opaque "Classification:/Execution:/Task support:" block. The operational facts earn their place; the presentation instructions are padding in a tool definition.

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

Completeness2/5

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

An output schema exists, so return values need not be described, and the mutation semantics are covered reasonably. However, for a 25-parameter nested mutation with billable cost and no parameter documentation, the definition is far too thin — an agent cannot construct a valid premium or image-editor-sourced call from this text alone.

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

Parameters2/5

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

With 25 parameters at 0% schema description coverage, the description explains only idempotencyKey (derivation behavior) and seatId/seatName (required when multiple seats exist). Critical inputs — engine, qualityMode, premiumBudgetUsd (capped at 100, cost-bearing), blankEditorSource/imageEditorSource digests, reassessVersionId, and the market/locale family — get no meaning at all, leaving the agent to guess from bare property names.

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

Purpose3/5

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

The description supplies a verb chain ("Design, render and save") and a resource ("working version"), but the object is buried in jargon: "review-held editable scratch Display working version" is hard to map to a concrete artifact. It never contrasts with near-neighbours like build_ad, create_creative_version, clone_creative, or revise_creative, so an agent cannot confidently pick it out of the 100+ siblings.

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?

Useful conditional guidance exists for preconditions: seatId or seatName must be supplied on the exact call when more than one seat is authorised, and mutation modes validate_only/preview_only/execute imply a dry-run option. But there is no statement of when to choose this over build_ad, create_creative_version, or clone_creative, which is the primary ambiguity given the dense sibling set.

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

All LeanrIQ tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?