# levea-mcp-server [Health: Active]

**Category:** 🧠 Knowledge & Memory  
**Repository:** https://github.com/brajendrak00068/openclaw-ai-video-editor  
**GitHub Stars:** 4  
**npm Downloads (last month):** 1837  
**Views:** 1  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/levea-mcp-server

## Description
Autonomous AI video editor over MCP: viral clips, captions, reframe, chroma key, MP4 export

## Tools
Capabilities this server exposes over MCP:

- **autonomous_edit** — PRIMARY EDIT TOOL. Run a natural-language video edit end-to-end. The backend agent plans, executes, verifies, and exports. Accepts anything from a one-liner ("generate 3 viral clips with captions and export") to a multi-paragraph creative brief. Returns the final mp4 URL synchronously when an export is queued. For long renders you may prefer queue_edit + check_job_status.
- **autonomous_edit_streaming** — Same as autonomous_edit but consumes the SSE event stream from the backend. Returns the final terminal result; per-step events (status, thinking, tool_call, tool_result, background_job_completed) are emitted as MCP progress notifications when the client supports them. Use for long-running briefs where users expect live progress.
- **queue_edit** — Fire-and-forget async path. Returns immediately with a `taskId`. Use when the edit may exceed the MCP request timeout (long viral-clip generation, multi-format export). Poll with check_task_status or get_active_task.
- **check_job_status** — Poll an export / async job by jobId (numeric). Supports wait:true to block until terminal.
- **check_task_status** — Poll a queued edit task by taskId (returned by queue_edit). Separate from job_id (exports).
- **get_active_task** — For a given project, return the currently-running edit task (if any). Useful for reconnecting MCP clients.
- **list_caption_templates** — List the user's caption templates (builtin + saved). Returns slugs, display names, and style summaries.
- **apply_caption_template** — Apply a caption template to a scene synchronously (without going through autonomous_edit).
- **save_caption_template** — Save a new caption template definition the user can re-apply later.
- **save_current_caption_template** — Snapshot the caption styling currently applied to a project / scene as a reusable template.
- **delete_caption_template** — Delete a user-saved caption template by id.
- **list_brand_kits** — List the user's brand profiles (palette, fonts, logo, voice, gradeBias). Returns ids you pass to autonomous_edit's brandId / projectBrandId.
- **get_brand_kit** — Fetch one brand profile by id.
- **create_brand_kit** — Create a brand profile. Fields: name, palette {primary,secondary,accent,neutral,extras[]}, fonts {primary,secondary}, gradeBias (warm/cool/neutral/vivid/muted), forbiddenColors[], logoAssetUrl, voiceTone, enforcement (soft|hard).
- **update_brand_kit** — Update an existing brand profile by id. Body has the same shape as create_brand_kit.
- **delete_brand_kit** — Delete a brand profile.
- **list_projects** — List the user's projects.
- **get_project** — Fetch one project by id.
- **create_project** — Create a new project. Optional initial brandId.
- **delete_project** — Delete a project.
- **asset_upload_url** — Get a signed upload URL for a local file (so users do not have to host video / image / audio elsewhere before passing video_url to autonomous_edit). Body: { fileName, fileType?, contentType?, projectId? }. Returns { uploadUrl, fileUrl }; PUT the file bytes to uploadUrl, then pass fileUrl to autonomous_edit as video_url / in assets[].
- **list_assets** — List the user's uploaded assets. Optional projectId filter.
- **delete_asset** — Delete an asset by file name (the path returned by asset_upload_url).
- **transcribe_asset** — Force transcription of an existing asset. Returns transcript words + timing.
- **editor_health** — Connectivity diagnostic against LEVEA_API_URL. Unauthenticated.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "levea-mcp-server": {
    "command": "npx",
    "args": ["-y","levea-mcp-server"],
    "env": {
      "LEVEA_API_KEY": ""
    }
  }
}
```

**Requires environment variables:** `LEVEA_API_KEY` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What levea-mcp-server does

Levea turns natural-language editing instructions into structured video projects and rendered outputs through MCP. An agent can submit a short request or a detailed creative brief for operations such as cutting viral clips, adding captions, reframing footage, applying chroma key effects, and exporting MP4 video.

The server also exposes project and media-management functions. Agents can create and inspect projects, upload source files through signed URLs, list or delete assets, and request transcription with word-level timing. Caption templates and brand kits can be created, reused, updated, or removed independently of a full autonomous edit.

## How it works

The local MCP wrapper sends editing requests to Levea’s hosted production engine using an API key. Levea separates creative planning from media execution: a planning layer interprets the prompt, source media, transcripts, and timeline requirements, then produces typed editing operations for the production harness.

Projects are represented as structured scene data containing timelines, tracks, layers, assets, text, masks, audio, effects, and brand settings. The hosted engine performs the media work on cloud GPU infrastructure, including transcription, speaker-aware reframing, caption rendering, motion graphics, audio processing, and MP4 export. Edits can return a final result synchronously, emit progress events through the streaming tool, or run as queued tasks for longer renders.

Use `check_job_status` for export jobs, `check_task_status` for queued edit tasks, and `get_active_task` to reconnect to an edit already running for a project.

## Setup and configuration

Create a Levea API key through Livecore and provide it to the MCP client configuration. The server is intended for MCP-compatible editors, including Claude Desktop, Claude Code, Cursor, Cline, and Windsurf.

Claude Code can install the associated plugin with:

```text
/plugin install levea-video-editor --marketplace brajendrak00068/agentic-ai-video-production
```

After installation, select `levea` from `/mcp` and authenticate in the browser. For direct MCP use, register the `levea-mcp-server` stdio wrapper in the client and supply the API key. Local media can be uploaded by first requesting a signed URL with `asset_upload_url`, sending the file bytes to that URL, and passing the returned file URL to an edit request.

## Tools and capabilities

- Submit autonomous editing briefs synchronously or through streamed progress notifications.
- Queue long-running edits and poll task or export status.
- List, apply, save, snapshot, and delete caption templates.
- Create and manage brand kits with palettes, fonts, logos, tone, grading bias, and enforcement settings.
- Create, inspect, list, and delete projects and their uploaded assets.
- Request signed uploads and transcribe existing media with timed transcript words.
- Run an unauthenticated connectivity check against the configured Levea API endpoint.

## Limitations and notes

The MCP wrapper is open source under the MIT license, while video production runs through Levea’s hosted service. An API key is required for authenticated editing operations. Heavy rendering, transcription, tracking, motion graphics, and export execute remotely rather than on the client machine.

Synchronous autonomous edits return the final MP4 URL when an export is queued, but longer work may outlast an MCP request timeout. In those cases, use `queue_edit`, the streaming variant, or the task and job status tools. The health diagnostic does not require authentication.

_Full upstream README: https://allmcps.com/mcp/levea-mcp-server/readme_