Levea Agentic Video Editor for OpenClaw

Connect OpenClaw agents to the Levea video editor API. The skill supports one primary free-form tool, autonomous_edit, plus a deterministic allowlist for callers that already know the exact structured edit to run.

What Users Can Ask

"Make this clip vertical for TikTok, remove silences, add bold captions, and export."
"Turn this long video into five 15-30 second viral clips."
"Key out the green screen, add a studio background, and keep the speaker centered."
"Add B-roll over the product mention and duck the music under speech."
"Export this project as an MP4."

Search Terms This Skill Covers

This listing is intentionally written for the phrases OpenClaw and npm users are likely to search:

OpenClaw AI video editor, OpenClaw video editing skill, ClawHub video editor
natural-language video editing, agentic video editing, AI video automation
viral clip generator, short-form video clips, TikTok video editor, Instagram Reels editor, YouTube Shorts editor
auto captions, caption generator, subtitles, video captions, social media captions
vertical video, 9:16 reframe, multi-platform export, MP4 export
chroma key, green screen removal, background removal, video background replacement
B-roll generation, motion tracking, audio cleanup, silence removal, voiceover, background music

Recommended ClawHub publish tags:

clawhub skill publish . --tags latest,video,ai-video-editor,openclaw-skill,viral-clips,captions,vertical-video,chroma-key,audio-cleanup,export

API Surface

Base URL: {ADSCENE_API_URL}

For production Livecore, set:

export ADSCENE_API_URL="https://api.livecore.ai"

The full OpenClaw execute URL is:

https://api.livecore.ai/api/v1/misc/openclaw/v1/execute

Use https://studio.livecore.ai/ to sign up, log in, and generate an OpenClaw API key from the editor account UI.

Do not set ADSCENE_API_URL to https://studio.livecore.ai or to the in-product editor route /api/v1/misc/editor/. Studio is the user-facing app; OpenClaw requests should go to the API-key route below on https://api.livecore.ai.

Endpoint	Purpose
`POST /api/v1/misc/openclaw/v1/execute`	Run `autonomous_edit` or a deterministic tool
`GET /api/v1/misc/openclaw/v1/jobs/{jobId}`	Poll async render or generation jobs
`GET /api/v1/misc/openclaw/v1/tools`	List available tool names
`GET /api/v1/misc/openclaw/v1/health`	Health check

Auth:

Authorization: Bearer {ADSCENE_API_KEY}

Primary Tool

Use autonomous_edit for natural-language video edits. The backend classifies intent, plans canonical edit steps, executes through the same safety gates as the in-product editor, verifies the result, and queues export when needed.

curl -sS -X POST "$ADSCENE_API_URL/api/v1/misc/openclaw/v1/execute" \
  -H "Authorization: Bearer $ADSCENE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "tool": "autonomous_edit",
    "params": {
      "prompt": "Make this a TikTok-ready viral clip: vertical reframe, add bold captions, remove silences, and export."
    },
    "project_id": "my-project"
  }'

Deterministic Tools

Use these when the caller already has structured parameters and wants to skip intent planning.

Tool	Canonical action	Use when
`read_scene`	`READ_SCENE`	Inspect the current timeline
`read_media`	`QUERY_ASSETS`	List gallery assets
`read_visual`	`READ_VISUAL`	Analyze frames
`query_transcript`	`QUERY_TRANSCRIPT`	Search transcript text or timestamps
`scene_update`	`UPDATE_LAYER`	Mutate known layer properties
`scene_insert`	`CREATE_LAYER`	Add a video, audio, text, image, shape, or adjustment layer
`scene_timing`	`SCENE_TIMING`	Trim, retime, or reposition a layer
`scene_mask`	`APPLY_MASK`	Apply chroma, luma, alpha, or depth masking
`chroma_key`	`APPLY_MASK`	Green-screen or blue-screen keying
`split_screen`	`SPLIT_SCREEN`	Top-bottom, left-right, grid, or picture-in-picture layout
`caption_compose`	`GENERATE_CAPTIONS`	Generate captions from transcript
`media_treat`	`COLOR_GRADE`	Apply color treatment
`scene_track`	`TRACK_MOTION`	Face or object tracking
`clean_audio`	`CLEAN_AUDIO`	Remove silence, breaths, or filler words
`audio_mix` / `audio_mixing`	`AUDIO_MIXING`	Ducking, normalize, denoise, or EQ
`voiceover_add`	`GENERATE_VOICEOVER`	Generate voiceover
`music_generate`	`GENERATE_MUSIC`	Generate background music
`export_video`	`EXPORT_VIDEO`	Render MP4 output

Unknown tool names return UNKNOWN_TOOL with a hint to use autonomous_edit.

Response Expectations

JSON responses include:

success, status, and message
scene when an edit changed the project
videoUrl when export output is already available
jobId, activeTasks, or pendingAsyncJobs for async render/generation work
workingMemory when a plan is awaiting approval
verificationPassed and verificationIssues when verification ran

The execute endpoint also supports SSE with Accept: text/event-stream or ?stream=true.

Configuration

Set the API URL and key through OpenClaw skill configuration or the process environment. Create the key from the Studio app at https://studio.livecore.ai/, then use https://api.livecore.ai as the API URL. Keep keys out of code and shared logs.

{
  "skills": {
    "entries": {
      "openclaw_ai_video_editor": {
        "enabled": true,
        "env": {
          "ADSCENE_API_URL": "https://api.livecore.ai",
          "ADSCENE_API_KEY": "your-openclaw-api-key"
        }
      }
    }
  }
}

Safety Model

API-key authentication is required for execute and job polling.
Requests are rate-limited per API key.
Destructive actions require explicit confirmation parameters.
Mutating tools auto-queue an export only when the scene actually changed and no export is already queued.
Read-only tools never auto-export.

Supported Media

Video input: MP4, MOV, WebM, gallery IDs, HTTP/HTTPS URLs, or supported source URLs
Image input: JPG, PNG, WebP
Audio input: MP3, WAV, M4A, AAC
Output: MP4 for export, ZIP for multi-clip or multi-platform bundles

Support

For setup help, integration questions, or issue reports, contact:

brajendrak00068@gmail.com

OpenClaw AI Video Editor