# GmAnimato

> GmAnimato (https://animato.gmgard.moe, also https://animato.gmgard.us) generates short AI videos from images, reference clips and prompts (image-to-video, first/last-frame, reference-to-video). These static docs are public. The API needs a personal access token that the user creates on their Profile page.

## Skill
- [Agent skill (step-by-step: auth, upload, prompt, submit, poll, download, reuse)](/skill.md): the file an AI agent should load first.
- [Skill reference (endpoint matrix, field tables, limits, errors)](/skill-reference.md): details the skill points to.
- [Everything in one file](/llms-full.txt): skill.md + skill-reference.md concatenated for single-fetch agents.

## API
All calls send `Authorization: Bearer anim_pat_…`.
- [GET /api/auth/session](/api/auth/session): who am I and when the token expires.
- [GET /api/quota](/api/quota): daily credits used / limit / remaining.
- [GET /api/workflows](/api/workflows): available models with allowed steps, durations, resolutions, required LoRAs and cost.
- [GET /api/videos](/api/videos): your videos, paged.
- [POST /api/uploads](/api/uploads): create a draft from images / reference clips (i2v, fl2v, ref2v).
- [POST /api/i2v/kickoff](/api/i2v/kickoff): submit a draft for generation.
- [POST /api/video/update](/api/video/update): save the prompt of a draft.
- [POST /api/video/reuse](/api/video/reuse): create a new draft from an existing video's inputs.
- [POST /api/video/resolve-ref](/api/video/resolve-ref): turn a finished video id/URL into a reusable reference-video URL.
- Per-video routes (need an id, so not linked): `GET /api/video/{id}`, `GET /api/video/{id}/status`, `POST /api/video/{id}/reopen`, `POST /api/video/{id}/retry`, `GET /api/video/{id}/download?format=gif|webp`, `GET /media/{bucket}/{file}` (video file).

## Pages
- [Create a video](/new): the web UI the API mirrors.
- [My videos](/videos): list of your generations.
- [Profile](/profile): create and revoke API tokens.
- [Gallery](/gallery): published videos.