# Opus Video Maker

> Turns a brief, plus optional images, audio, video, PDFs or links, into a finished MP4: an AI director plans the shots, writes the animation as code, renders it frame by frame and adds voice-over. A video takes about 10-60 minutes and is billed to the account balance.

Agents connect over MCP (https://maker.vcorp.ai/mcp) or REST (https://maker.vcorp.ai/v1). Accounts are created by the site's admin; a user then connects an agent with OAuth (paste the MCP URL into Claude.ai, ChatGPT, Cursor or VS Code and log in) or with an API key made under "接入 Agent" in the web app.

Because a video takes long, nothing blocks until it is done: create a job, get a job_id at once, then long-poll with wait_for_video, check back later with get_video, or pass a callback_url and get a signed webhook.

## Docs

- [Agent guide](https://maker.vcorp.ai/docs/agents.md): setup for each client, every tool, the job lifecycle, webhooks, errors
- [OpenAPI](https://maker.vcorp.ai/openapi.json): the REST API, same operations as the MCP tools
- [Agent Skill](https://maker.vcorp.ai/agents/SKILL.md): drop into Claude Code or another agent to drive the workflow
- [MCP Server Card](https://maker.vcorp.ai/mcp/server-card)

## Connect

- MCP endpoint: https://maker.vcorp.ai/mcp (Streamable HTTP, stateless)
- Auth: OAuth 2.1 with dynamic client registration (metadata at https://maker.vcorp.ai/.well-known/oauth-protected-resource/mcp), or `Authorization: Bearer ovm_...`
- Claude Code: `claude mcp add --transport http opus-video https://maker.vcorp.ai/mcp --header "Authorization: Bearer ovm_..."`

## Optional

- [Web app](https://maker.vcorp.ai/)