Skip to main content
Glama

OpenClip

by app.openclip
OverviewDiscussionsAdmin

Generate Ugc Tool

generate_ugc

Generate a short (~7-12s) vertical UGC-style video clip from a creative brief. The service builds a start frame, renders an image-to-video clip, and returns it.

Account-gated. Free accounts (no subscription) get ONE generation per day (a teaser); subscribe to generate more, metered by the seconds you render. Paid accounts are credit-metered: a render is charged from your credit balance based on the seconds of video produced, settled when it finishes.

Provide the identity EITHER by a creator slug (discover them with list_ugc_creators; a saved spokesperson that bakes the brief, voice, locked seed, and start frame) OR a brief_name (a server-side brief preset) OR an inline brief object. Inline fields override a creator's baked values. A brief is NOT a prose prompt: it is a small set of creative fields the service composes into the start-frame image and the spoken delivery. Use these fields (all strings, all optional but more is better):

  • character: who is on camera (age, look, build, vibe; describe real texture, not model-polished)

  • wardrobe: what they wear

  • setting: where they are + the lighting/mood

  • framing: shot + camera placement (e.g. "phone propped on a table, medium-close")

  • vibe: energy and how they address the viewer

  • delivery: pacing of the line (e.g. "fast exasperated hook, then relaxed")

  • accent: spoken accent (e.g. "natural American accent")

  • sound: ambience under the voice

  • script: the spoken line (~24-30 words, ONE hook in the first 2s). This drives BOTH the audio AND the caption text, so write the exact words you want said + shown.

Example brief: {"character":"a woman in her late 20s, natural skin, tired but sharp", "wardrobe":"plain dark crewneck","setting":"a bright cafe, soft window light", "framing":"phone propped on the table, medium-close","vibe":"candid, talking to a mate","delivery":"fast hook then relaxed","accent":"natural American accent", "script":"i run a company and have no time to edit. now i drop one video in and get a week of clips back, captions and all. i just post."}

The clip length follows the script: it is sized to the words at a natural speaking pace, up to ~20 seconds (one LTX pass), so keep a script to roughly one clip. Optionally set top-level script, seed for reproducibility, ref (hashid of an uploaded image from create_upload to use as the start frame; omit to have one generated), and caption_preset to auto-caption the finished clip (one-shot).

Processing is ASYNCHRONOUS: this returns a ugc_job hashid and status: "queued". Poll get_ugc_job_status with that ugc_job until it reports completed. It then returns the downloadable video_url; once captioning is ready it also returns video

  • viral_moment ids you can pass to render_clip with a caption_preset. The nudge field is an optional suggestion.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
refNoOptional hashid of an uploaded image (from create_upload) to use as the start frame.
seedNoOptional seed for a reproducible render.
briefNoAn inline brief object. Use this OR `brief_name`.
voiceNoOptional voice as {accent, tone}, folded into the brief the worker speaks (overrides a creator default).
scriptNoThe spoken script. The clip length is sized to this at a natural pace (up to ~20 seconds), so keep it to roughly one clip.
creatorNoOptional creator slug from list_ugc_creators. Supplies a baked identity (brief, voice, locked seed, canonical start frame); inline fields override it.
brief_nameNoA server-side brief preset name (briefs/<name>.json). Use this OR inline `brief`.
start_frameNoOptional curated start-frame scene id (the identity anchor to open on); discover ids with list_ugc_creators / the studio.
caption_presetNoOptional caption preset key. When set, the finished clip is automatically captioned + rendered (one-shot); omit to get the raw clip and caption later via render_clip.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Added

TDQS

A4.9/5.0
Behavior5/5

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

With no annotations, the description carries the full burden, and it does so exceptionally. It discloses account gating, free-tier daily limits, credit metering, asynchronous processing with a job hashid, polling expectations, the relationship between script length and clip duration, override behavior for inline fields, and the one-shot caption path. There is no contradiction with annotations, which are empty.

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 long but organized into purposeful sections: what it does, account constraints, identity selection, brief field definitions, optional parameters, and async return behavior. It is front-loaded with the core purpose and the example brief is valuable. It loses one point because a few trailing details (e.g., 'The `nudge` field is an optional suggestion') add limited value and some sentences could be tightened without loss.

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?

There is no output schema, so the description correctly explains return semantics: it returns a `ugc_job` hashid with `status: 'queued'`, to be polled with get_ugc_job_status until `completed`, then yields `video_url` and later `video`/`viral_moment` ids. It also covers the full parameter space, including optional start_frame and voice, and the async workflow is fully specified. An agent has everything needed to invoke and follow up correctly.

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?

Although schema coverage is 100%, the description adds substantial meaning beyond the schema. It defines the inline brief's nine creative fields, gives a full example brief, explains how `creator`, `brief_name`, and `brief` interact with override precedence, clarifies how `script` drives both audio and captions, and explains `ref`, `seed`, and `caption_preset` behavior. This is far above the baseline for a schema-covered tool.

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 opens with a specific verb and resource: 'Generate a short (~7-12s) vertical UGC-style video clip from a creative brief.' It immediately distinguishes the tool from siblings like render_clip, edit_video, and list_ugc_creators by naming the workflow (start frame -> image-to-video -> clip). This is far beyond a tautology and gives the agent a precise model of what the tool produces.

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 description gives explicit selection guidance: use a `creator` slug, `brief_name`, or inline `brief`; poll get_ugc_job_status after the async call; use render_clip with caption_preset when captioning is ready; and omit `ref` to generate a start frame. It also explains account-gating and quotas, so the agent knows when generation is allowed. This is excellent when-to-use and when-to-use-alternative guidance.

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

All OpenClip toolsInspect
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