Metadata MCP Connector

Generate Flexible Brand Creative

generate_flexible_brand_creative

Generate an EDITABLE, layered brand creative: a clean AI background scene (no baked text) plus AI-placed, editable text/CTA/logo layers. Returns a structured layout document (JSON), NOT a flat image — the UI renders it in an editor the user can rearrange before saving.

USE FOR:

The DEFAULT for a standalone creative the user will review, refine, or use — "make me a creative", "design an ad", "give me options", "a creative I can tweak/edit".
Any request where the user should be able to move/retype the headline, restyle the CTA, swap the logo, add layers, or regenerate the background before finalizing.

DIFFERENT FROM generate_brand_creative: PREFER THIS for standalone creative requests. generate_brand_creative bakes headline + CTA into the pixels and returns one finished PNG; use that flat tool only when building an ad / feeding a campaign inline (the flat image flows straight into an ad, with no human edit-and-save step) or when a finished flat image is explicitly requested.

COPY DISCIPLINE: fewer layers = a cleaner layout the user can actually edit. Default to headline + cta (add subheadline only when it genuinely earns its place); pass body / disclaimer ONLY when the user explicitly asked for that copy.

COPY LIMITS (image copy is shorter than ad text): headline at most 45 characters and 8 words on a 1:1, 4:5 or 9:16 canvas (55 characters and 10 words on 16:9); cta at most 3 words and 18 characters; subheadline at most 90 characters. Copy over a limit is rejected with the limit named: shorten it and call again rather than splitting it across layers.

RETURNS: a layered document — canvas dimensions, background image URL, brand colors/logo, and an ordered list of layers (headline, subheadline, body, cta, disclaimer, logo) with normalized positions and styling.

Input Schema

TableJSON Schema

Name	Required	Description	Default
`cta`	No	Optional call-to-action button text (e.g. 'Learn more'): at most 3 words and 18 characters so the button stays on one line.
`body`	No	Optional supporting body copy. Pass ONLY when the user explicitly asked for body copy — extra layers crowd the layout.
`domain`	Yes	Company domain for brand DNA (e.g. 'metadata.io'). No protocol.
`headline`	No	Headline copy to place as an editable layer. Image copy, not ad text: at most 45 characters and 8 words on a 1:1, 4:5 or 9:16 canvas, 55 characters and 10 words on 16:9. Longer copy is rejected with the limit named, so it can be shortened and resent.
`platform`	No	Target platform (sets the default aspect ratio). Default: 'linkedin'.	linkedin
`disclaimer`	No	Optional fine-print/disclaimer copy. Pass ONLY when the user explicitly asked for a disclaimer.
`subheadline`	No	Optional subheadline copy, at most 90 characters; pass it only when it adds a fact the headline lacks.
`aspect_ratio`	No	Canvas shape, overriding the one 'platform' implies. Pass this whenever the user names a size or shape: '1:1' for square (e.g. 1080x1080), '16:9' for wide/landscape, '9:16' for a full-height story, '4:5' for portrait. Omit to keep the platform's default shape.
`include_logo`	No	Add a brand logo layer (resolved from the brand kit). Default: true.
`instructions`	Yes	Natural-language direction for the background scene and overall creative (e.g. 'clean B2B scene for an AI analytics launch').

Schema Changelog

Changes observed during successful MCP inspections.

Changed6 schema fields changedSep 17, 2026

changedInput schema / properties / cta / description

Previous value: -"Optional call-to-action button text (e.g. 'Learn more')."New value: +"Optional call-to-action button text (e.g. 'Learn more'): at most 3 words and 18 characters so the button stays on one line."

addedInput schema / properties / cta / maxLength
```
Added value: +18
```

changedInput schema / properties / headline / description

Previous value: -"Headline copy to place as an editable layer."New value: +"Headline copy to place as an editable layer. Image copy, not ad text: at most 45 characters and 8 words on a 1:1, 4:5 or 9:16 canvas, 55 characters and 10 words on 16:9. Longer copy is rejected with the limit named, so it can be shortened and resent."

addedInput schema / properties / headline / maxLength
```
Added value: +55
```

changedInput schema / properties / subheadline / description

Previous value: -"Optional subheadline copy."New value: +"Optional subheadline copy, at most 90 characters; pass it only when it adds a fact the headline lacks."

addedInput schema / properties / subheadline / maxLength
```
Added value: +90
```

First observedSep 16, 2026

TDQS

A4.9/5.0

Behavior5/5

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

Beyond the annotations, the description discloses key behavioral traits: it returns an editable layered document rather than a finished image, copy over the limits is rejected with the limit named, and body/disclaimer should only be passed when explicitly requested. These details materially change what the agent should expect and how it should react to failures.

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

Conciseness5/5

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

The description is long but extremely well structured with clear sections: main definition, USE FOR, DIFFERENT FROM, COPY DISCIPLINE, COPY LIMITS, and RETURNS. Every section earns its place for a tool with 10 parameters and no output schema, and the core distinction is front-loaded.

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?

Given the tool's complexity and absence of an output schema, the description is complete: it explains the returned structure (canvas, background image URL, brand colors/logo, ordered layers), how the UI consumes it, default platform behavior, and copy rejection rules. An agent has enough context to invoke the tool correctly and set expectations.

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 already 100%, so the baseline is 3, but the description adds meaningful cross-parameter semantics: copy-length limits vary by aspect ratio (45 chars on 1:1, 4:5, 9:16 vs 55 on 16:9), and it reinforces which copy fields are optional and when to include them. This goes beyond the individual schema descriptions, though some redundancy remains.

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 names a specific verb and resource — 'Generate an EDITABLE, layered brand creative' — and clearly states the output is a structured JSON layout, not a flat image. It also distinguishes itself from sibling generate_brand_creative by explaining exactly what that alternative returns instead.

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?

The 'USE FOR' section explicitly lists when this tool is the default, and the 'DIFFERENT FROM' section names generate_brand_creative and explains when to prefer that sibling. This gives an agent explicit, actionable routing criteria with no inference needed.

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

All Metadata MCP Connector 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?