Skip to main content
Glama

Generate an AI motion clip

autoshort_generate_motion_clip
Idempotent

Generate a single silent AI motion clip from a pasted text-to-video prompt, preserving the prompt exactly. Use it for cinematic scene, camera, and lighting prompts rather than narrated videos.

Instructions

Turn a text-to-video prompt into ONE AI motion clip (Seedance 2.0). The prompt is sent EXACTLY as written: no script is written, no still images are made, no voice is added, nothing is rewritten or shortened. USE THIS TOOL when the user pastes or asks for a cinematic / video-model prompt (camera moves, lens, lighting, a scene description, "9:16 vertical", "slow dolly-in", the kind ChatGPT or Claude writes for Seedance, Sora or Veo). For a narrated story or facts video use autoshort_generate_chat_video. The clip is silent (no audio, no lip-sync, no talking), 4-12 seconds (default 5), resolution 480p (default 480p; higher resolutions need a higher plan: PLAN_TIER_LOCKED). Price: 20 credits per second, so 5 s = 100 credits and 12 s = 240; call autoshort_quote_video first and tell the user the price before creating. The credits are held when the clip is accepted and returned in full if no clip comes back. The only text added to the prompt is a short server suffix ("No spoken dialogue, no lip-sync, no talking to camera, no watermark."); the answer echoes promptSent and the seed so the user can see and repeat exactly what was made. Not accepted (CONTENT_POLICY, nothing charged): real or famous people, minors, sexual content, protected characters or brands, deepfakes and lip-sync. A first-frame photo is not supported yet. Optional character_id: the id of one of YOUR character sheets (autoshort_get_character_sheet); the clip then starts from that sheet's chosen image, so the first frame shows your character. The character can still drift while it moves; it is not a guarantee of an identical face. A sheet that is not yours or not ready is CHARACTER_NOT_FOUND (nothing charged). Same price. Optional templateId: one of the 26 ready clip templates from autoshort_list_templates (kind "motion"); its prompt, seconds and aspect ratio are used. If you also pass prompt, the prompt wins and the template is ignored. An unknown templateId is refused (INVALID_INPUT, nothing charged). Trial accounts get up to 10 clips of up to 5 s each (MOTION_TRIAL_LIMIT); at most 2 clips run at once; a daily capacity limit applies (DAILY_CAP_REACHED). Rendering is asynchronous: poll autoshort_get_video every 20 seconds until COMPLETED (videoUrl is an MP4) or FAILED (errorCode, creditsRefunded). Pass idempotencyKey so a retry never charges twice.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "autoshort_generate_motion_clip"
  }
}

Input Schema

NameRequiredDescriptionDefault
seedNoOptional. The answer always returns the seed used; pass it again to ask for a similar take (the provider treats it as best effort).
promptNoThe video prompt, exactly as the user wrote or pasted it (8-2000 characters). Do not summarise, translate or "improve" it: it is sent to the video model unchanged. Required unless templateId is given.
durationNoClip length in whole seconds (default 5). Trial accounts: at most 5.
resolutionNoOutput resolution (default 480p). Higher ones cost more credits per second and need a higher plan.
templateIdNoOptional. The id of a motion clip template (autoshort_list_templates, kind "motion"): its ready-made prompt is sent as written. Ignored when prompt is also given.
aspect_ratioNoFrame shape (default 9:16, vertical).
character_idNoOptional. The id of one of your own READY character sheets: the clip starts from that sheet's chosen image. Describe the motion in the prompt.
idempotencyKeyNoRetry-safe: a repeat with the same key returns the first answer (replayed: true) and charges nothing.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv2.3.0

TDQS

A4.9/5.0
Behavior5/5

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

Annotations only declare non-readonly, non-destructive, idempotent, open-world; the description goes far beyond that with pricing (20 credits/second with worked examples), hold-and-refund semantics, the server-appended suffix text, content policy rejections that cost nothing, trial caps, concurrency and daily limits, and asynchronous completion states with errorCode/creditsRefunded. It also honestly discloses a limitation of character_id ('the character can still drift... not a guarantee of an identical face').

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 purpose, the use-when trigger, and the alternative are front-loaded before pricing and edge cases, so an agent can act on the first two sentences. The remainder is dense and heavily capitalized (EXACTLY, ONE, USE THIS TOOL), and the pricing arithmetic and trial-limit details could be tightened without losing 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?

With eight optional parameters, no output schema, and an async render, the description supplies everything an agent needs: cost, refund behavior, the required quote-then-generate-then-poll sequence, the shape of the polled result (videoUrl MP4, errorCode, creditsRefunded), and idempotency guidance. Nothing material is left to guesswork.

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?

Schema coverage is 100%, so the baseline would be 3, but the description adds genuine meaning the schema does not: templateId is ignored when prompt is present, character_id starts the clip from the sheet's chosen image with a drift caveat, seed enables repeat takes, and unknown templateId/foreign sheet produce specific error codes with no charge. This is precedence and failure-mode semantics, not restatement.

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?

Opens with a specific verb and resource ('Turn a text-to-video prompt into ONE AI motion clip') and immediately bounds scope with negatives (no script, no stills, no voice, nothing rewritten). It also explicitly names the sibling to use instead for narrated content (autoshort_generate_chat_video), so an agent can separate it from the other generate_* tools without opening a schema.

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?

Gives an explicit trigger ('USE THIS TOOL when the user pastes or asks for a cinematic / video-model prompt') with concrete examples of such prompts, names the alternative for narrated story/facts videos, and prescribes the surrounding workflow (call autoshort_quote_video first, poll autoshort_get_video every 20 seconds). Exclusions are stated, not inferred.

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