Skip to main content
Glama

Animate one image into a clip

autoshort_generate_image_to_video
Idempotent

Animates a single reference image into a silent 5-second AI video, using the photo as the first frame. Suits products, pets, scenery, food and illustrations; realistic people are rejected.

Instructions

Animate ONE reference image into a silent 5-second AI-generated video (for a multi-scene video built from several of the user's pictures, use image_urls on autoshort_generate_chat_video instead). Send the image as image_url (a public https link, JPEG/PNG/WebP, at most 10 MB) or as imageBase64; the stdio package also accepts a local imagePath and uploads it. The photo becomes the first frame of the clip. ACCEPTED: products, pets, scenery, food, illustrations, and cartoon, anime or other clearly stylised characters (not realistic people, whether real or AI-generated). REJECTED before anything is charged: photos of real people (any realistic human face, to prevent deepfakes), sexual content, anything that may show a child; the answer says why. Only send images the user owns or may use. Costs 100 credits (20 credits per second x 5 s), charged the moment the job is accepted (not on delivery). Any plan with enough credits; a never-paid account gets one AI clip until its first payment. Rendering is automatic (OpenRouter Seedance 2.0 Mini, 480p, no audio): ready in up to about 13 minutes, typically less (not yet measured), so poll autoshort_get_video every 20-30 seconds. If the clip cannot be made (for example the video provider refuses the image on content-policy grounds) the video ends FAILED with the specific reason and the credits are refunded automatically; if it does not come back, check autoshort_get_credit_balance and report a missing refund. The result is labelled AI-generated (aiGenerated: true); keep that label when posting. Retry-safe: the same image + motionPrompt returns the first job (duplicate: true) for 24 hours instead of charging again; pass idempotencyKey to force a separate clip from identical inputs. To turn a TEXT prompt into a moving clip automatically (no photo, no human operator), use autoshort_generate_motion_clip instead.

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": {
      "motionPrompt": "<motionPrompt>"
    },
    "name": "autoshort_generate_image_to_video"
  }
}

Input Schema

NameRequiredDescriptionDefault
imagePathNoAbsolute path to a local JPG, PNG or WebP file up to 10MB (read and uploaded by this local server). Use instead of image_url or imageBase64.
image_urlNoPublic https link to the JPG, PNG or WebP image (≤10MB). Alternative to imageBase64; send one of the two.
imageBase64NoThe JPG, PNG or WebP file (≤10MB), base64-encoded. Alternative to image_url.
motionPromptYesMotion, camera and scene direction.
imageFileNameNoOptional original file name.
idempotencyKeyNo

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 the safety profile (readOnly=false, idempotent=true, destructive=false, openWorld=true); the description adds the material behavior an agent cannot infer: 100-credit charge at job acceptance with refunds on provider-side failure, content-policy rejections decided before charging with a stated reason, ~13-minute rendering with polling advice (autoshort_get_video every 20-30 s), a 24-hour duplicate window with a force-new idempotencyKey, and a required aiGenerated label. This goes well beyond what the annotations convey.

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?

Front-loaded with purpose and the sibling alternative, then grouped logically into inputs, accept/reject policy, cost, rendering, failure/refund, and retry semantics. It is dense — roughly 280 words in a single block — so the cost, refund, and retry sections ask a lot of a reader, but no sentence is filler and each adds distinct operative detail.

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 an openWorld, credit-charging, asynchronous generation tool with no output schema, this covers everything an agent needs: input formats, content prohibitions, cost and charge timing, polling cadence, refund path including a balance check fallback, and duplicate handling. Nothing material is left to inference.

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 83%, yet the description still adds real meaning: it explains the mutually exclusive image input paths (image_url / imageBase64, plus imagePath on the stdio package) and what each costs, and clarifies idempotencyKey semantics — force a separate clip from identical inputs — which the schema leaves undocumented. It also states the image becomes the first frame, which is behavioral meaning not present in the parameter docs.

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 — animate ONE reference image into a silent 5-second AI video — and immediately names the sibling to use for multi-image scenes (image_urls on autoshort_generate_chat_video). The closing sentence distinguishes it from autoshort_generate_motion_clip by input type (photo vs. text prompt).

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?

Explicitly routes the agent: use this for a single image, use autoshort_generate_chat_video when building a multi-scene video from several pictures, and use autoshort_generate_motion_clip when there is no photo. It also states input prerequisites (public https, formats, 10 MB cap) and the eligibility constraint (any plan with enough credits, one free AI clip for never-paid accounts).

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