Skip to main content
Glama
unifically-dev

UnificAlly MCP Server

Official
by unifically-dev
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Hybrid

Generate video

generate_video
Destructive

Create videos from text prompts with optional start and end frames using AI models like Veo or Kling. Returns the video URL.

Instructions

Generate a video from a text prompt (and optional start/end frames). Defaults to Google Veo 3.1 Fast, which includes native audio. Switch model with the model parameter (Veo, Kling, Hailuo/MiniMax, and more). The tool waits for the result and returns the video URL; pass callback_url for long jobs to return immediately.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
seedNoOptional seed for reproducible output. Same seed + same input gives the same result on supported models.
waitNoDefault true: wait for the job and return the finished result. Set false to get the task_id back immediately and poll with get_task — use for long jobs (high-quality video, full songs) or when the connection might time out.
modelNoModel, as a short alias or full id. Aliases: 'veo' (default), 'veo-quality', 'veo-lite', 'kling', 'kling-omni', 'hailuo', 'seedance', 'grok-video', 'grok-1.5'. Full ids like 'google/veo-3.1-fast' also work. Call list_models for the catalog, and read a model's docs page for its exact parameters.
promptYesText prompt describing the desired video.
durationNoVideo length in seconds. Omit for the model's default. Allowed values are per-model.
image_urlNoOptional starting frame for image-to-video. This tool maps it to whatever field the chosen model wants, so pass it here rather than the model's own field name. Must be a public https URL — never a local path, never inline base64. For a local file, upload it first with a direct PUT to files.unifically.com/upload and pass the returned file_url.
aspect_ratioNoOutput aspect ratio, e.g. '16:9'. Per-model: some accept different sets, some have none.
callback_urlNoOptional webhook URL. When set, the tool returns immediately with the task_id and UnificAlly POSTs the result to this URL when the task finishes. Webhook body: { task_id, status: 'completed'|'failed', data: { video_url | image_url | audio_url | ... }, error_message? } — note the media arrives under 'data', not 'output'. Use for long jobs instead of waiting.
confirm_costNoOnly relevant if the operator enabled a spend guard. When a call exceeds the configured confirm threshold it returns a quote; retry with confirm_cost: true to proceed. Safe to ignore otherwise.
end_image_urlNoOptional final frame (start-frame + end-frame interpolation). Only some models support one. Must be a public https URL — never a local path, never inline base64. For a local file, upload it first with a direct PUT to files.unifically.com/upload and pass the returned file_url.
provider_optionsNoEscape hatch for model-specific fields not exposed as named parameters (e.g. Kling multi_shots, Flux steps). Merged into the API 'input' payload. Rejected: fields the tool already manages (prompt, media URLs, task_id, ...) and reserved routing/billing/auth keys (model, callback_url, dry_run, api_key, authorization) — switch models with the 'model' parameter, not here. See https://docs.unifically.com for per-model fields.
reference_image_urlsNoOptional reference images for style/subject — each must be a public https URL (upload local files first — never a path or base64). Support, the max count, and whether they combine with a start frame are per-model.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv1.1.0

TDQS

A4.3/5.0
Behavior4/5

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

Annotations already declare destructiveHint=true, openWorldHint=true and idempotentHint=false, so the safety profile is covered. The description goes beyond that by disclosing synchronous-by-default waiting behavior, the callback webhook payload shape (including the non-obvious 'data' key rather than 'output'), and the spend-guard/confirm_cost retry flow, which annotations do not convey.

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?

Three tightly packed sentences with zero filler: purpose first, then defaults and model switching, then return/wait semantics with the callback escape hatch. Front-loaded and each clause carries information.

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 generation tool with no output schema, the description still covers the result (video URL) and the async path (immediate task_id plus webhook payload). Combined with the exhaustive schema and annotations, an agent has everything needed to call it correctly.

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

Parameters3/5

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

Schema description coverage is 100%, so every parameter is already documented in the schema, including the model aliases and the callback body. The description largely restates that (default Veo 3.1 Fast, native audio, switch model with model), adding little new parameter-level syntax beyond the baseline.

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?

States a specific verb+resource ("Generate a video") and immediately scopes the input modality ("from a text prompt (and optional start/end frames)"), which cleanly separates it from edit_video, extend_video, and generate_image. An agent can identify the tool 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?

Gives concrete routing conditions: default model, how to switch models via the model parameter, and explicitly when to prefer callback_url ("for long jobs to return immediately"). It does not, however, contrast with sibling tools like edit_video or extend_video, so the when-not guidance is absent.

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

Deploy Server

Other Tools