video-editor-ai-agent

作者 Don-Uwe已验证

:four_leaf_clover: ai agent video editor to feed raw footage and a creative brief, and a coordinated agent ensemble handles shot selection, agentic video trimming, video rendeing and agentic ai qualify review

125
Stars
1,024
Forks
TypeScript
语言
2026/8/23
添加时间

⚠️ 第三方软件声明

本 Skill 为第三方开源软件,独立托管于 GitHub。SkillTip 仅为信息目录,不控制或维护底层仓库。所显示的安全检查为自动化且范围有限,安装前请自行审查源码。

阅读服务条款

安装

添加到你的 Claude Code skills 目录:

# Add to your Claude Code skills
git clone https://github.com/Don-Uwe/video-editor-ai-agent

快速入门

使用 video-editor-ai-agent 等 Skills 的指南。

安全报告

已验证

上次扫描:—

{
  "status": "PASSED",
  "issues": []
}

README.md

Agentic Video Editor

Production-oriented fork of an AI-assisted video pipeline. Feed raw footage and a creative brief; a coordinated agent ensemble handles shot selection, trimming, rendering, and quality review.

The CLI is the primary, fully supported interface. AVE Studio (Next.js + Hono API) provides an experimental non-linear editor shell with optional Redis-backed persistence.

Built as a TypeScript monorepo (apps/ + packages/) with shared infrastructure in @ave/core (config, logging, errors, Redis, Zod schemas).


Feature Highlights

CapabilityDescription
Agent pipelineDirector, Trim Refiner, Editor, and Reviewer orchestrated from YAML manifests
PreprocessingFootage indexing via ffprobe (scene detection / transcription are planned enhancements)
Retry loopReviewer-driven quality gate with configurable thresholds and versioned outputs
Style templatesStructured YAML guidance for pacing, overlays, and segment structure
AVE StudioWeb UI with timeline, monitors, inspector, and live job streaming
PersistenceOptional Redis layer for job snapshots and Studio cache
Configurable securityCORS and filesystem browse roots controlled via environment variables

Architecture

flowchart TB
    subgraph Input
        Footage[Raw footage folder]
        Brief[Creative brief JSON]
        Pipeline[Pipeline YAML]
    end

    subgraph Core["TypeScript core (@ave/domain)"]
        Pre[Preprocess\nffprobe indexing]
        Runner[Pipeline runner]
        Director[Director agent\nGemini]
        Trim[Trim Refiner]
        Editor[Editor / FFmpeg]
        Reviewer[Reviewer agent\nGemini]
    end

    subgraph Output
        MP4[Rendered MP4]
        Scores[Review scores]
    end

    Footage --> Pre
    Pre --> Runner
    Brief --> Runner
    Pipeline --> Runner
    Runner --> Director --> Trim --> Editor --> Reviewer
    Reviewer -->|score below threshold| Director
    Editor --> MP4
    Reviewer --> Scores

Web stack

flowchart LR
    Browser[Browser\nNext.js Studio]
    NextAPI[Next.js API routes\n/cache]
    Hono[Hono API\n@ave/api]
    Redis[(Redis\noptional)]
    Pipeline[Pipeline runner]

    Browser -->|REST + WebSocket| Hono
    Browser --> NextAPI
    NextAPI --> Redis
    Hono --> Redis
    Hono --> Pipeline

Installation

Prerequisites

Setup

git clone https://github.com/your-org/agentic-video-editor.git
cd agentic-video-editor

npm install
cp .env.example .env
# Edit GOOGLE_API_KEY and optional Redis settings

Environment variables are loaded automatically by the CLI, API, and Studio.


Configuration

VariableDefaultPurpose
GOOGLE_API_KEYRequired for Gemini agents
AVE_LOG_LEVELinfoLogging verbosity
AVE_OUTPUT_DIRoutputRender output directory
AVE_CORS_ORIGINShttp://localhost:3000,...Allowed browser origins
AVE_BROWSE_ROOTS~Comma-separated roots for /api/browse
REDIS_ENABLEDtrueToggle Redis features
REDIS_URLredis://127.0.0.1:6379Redis connection URL
REDIS_KEY_PREFIXave:Key namespace prefix
PORT8000Hono API listen port
NEXT_PUBLIC_API_URL``Override API base URL in Studio

See .env.example for the full list including Redis tuning options.


Usage (CLI)

npm run dev:cli -- edit \
  --footage-dir /path/to/footage \
  --brief '{"product": "My Product", "audience": "Women 25-45", "tone": "authentic", "duration_seconds": 30}' \
  --pipeline pipelines/ugc-ad.yaml \
  --style styles/dtc-testimonial.yaml

Briefs may be inline JSON or a path to a .json file. Outputs land in output/ with versioned filenames when the reviewer triggers retries.

Creative brief schema

{
  "product": "Product name",
  "audience": "Target demographic",
  "tone": "energetic, calm, professional",
  "duration_seconds": 30,
  "style_ref": "styles/dtc-testimonial.yaml"
}

Development

Run the CLI

npm run dev:cli -- edit --footage-dir ./footage --brief brief.json

Run AVE Studio

Terminal 1 — API:

npm run dev:api

Terminal 2 — Studio:

npm run dev:studio

Open http://localhost:3000

Quality commands

npm run validate          # typecheck + lint + test + build (all workspaces)
npm run typecheck
npm run lint
npm run test
npm run build

On Windows, scripts/validate.ps1 runs the same checks.


Testing

SuiteScope
npm test (root)Vitest unit tests under tests/unit/
Studio lintESLint + TypeScript in @ave/studio

Core pipeline integration tests are intentionally deferred — they require Gemini credentials and FFmpeg fixtures.


Project Structure

agentic-video-editor/
├── apps/
│   ├── api/                 # Hono REST + WebSocket (@ave/api)
│   └── studio/              # Next.js frontend (@ave/studio)
├── packages/
│   ├── core/                # Config, logging, errors, Redis, schemas
│   ├── domain/              # Pipeline, agents, FFmpeg tools
│   └── cli/                 # `ave` CLI entry point
├── pipelines/               # YAML pipeline manifests
├── styles/                  # Director style templates
├── tests/unit/              # Vitest unit tests
├── docs/internal/           # Maintainer audit notes
└── scripts/                 # validate.ps1 / validate.sh

Structure decisions are documented in docs/internal/STRUCTURE.md.


Troubleshooting

SymptomLikely causeFix
GOOGLE_API_KEY errorsMissing or invalid keySet in .env or export in shell
Browse returns 403Path outside AVE_BROWSE_ROOTSAdd parent directory to roots
Redis unavailableServer not runningStart Redis or set REDIS_ENABLED=false
FFmpeg not foundBinary not on PATHInstall FFmpeg and verify with ffmpeg -version
Studio cannot reach APIWrong proxy targetSet NEXT_PUBLIC_API_URL=http://localhost:8000

Check Redis connectivity from Studio:

curl http://localhost:3000/api/cache/status

FAQ

Is the web UI production-ready?
No. AVE Studio is experimental. Use the CLI for reliable workflows.

Do I need Redis?
No. The app runs without Redis; persistence and cache features degrade gracefully.

Can I add custom agents?
Implement an agent under packages/domain/src/agents/ and reference it in a pipeline YAML manifest.

How are retries versioned?
Each reviewer-triggered retry writes {name}_v{N}.mp4 so you can compare iterations.


Contributing

  1. Fork the repository and create a feature branch.
  2. Run npm run validate.
  3. Keep commits focused; include tests for behavioral changes.
  4. Open a pull request with a clear summary and test plan.

License

MIT

常见问题

What is video-editor-ai-agent?

video-editor-ai-agent is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by Don-Uwe. :four_leaf_clover: ai agent video editor to feed raw footage and a creative brief, and a coordinated agent ensemble handles shot selection, agentic video trimming, video rendeing and agentic ai qualify review. It has 125 GitHub stars.

Is video-editor-ai-agent safe to use?

Yes. video-editor-ai-agent passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.

How do I install video-editor-ai-agent?

Clone the repository with "git clone https://github.com/Don-Uwe/video-editor-ai-agent" and add it to your Claude Code skills directory (see the Installation section above).

What programming language is video-editor-ai-agent written in?

video-editor-ai-agent is primarily written in TypeScript. It is open-source under Don-Uwe on GitHub, so you can review or fork the full source.

Are there alternatives to video-editor-ai-agent?

Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh video-editor-ai-agent against similar tools.

评论 (0)

暂无评论,成为第一个分享想法的人!

ECC

by affaan-m

10

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

242,21936,702JavaScript
AI 智能体ai-agentsanthropicclaude-code
查看详情
15

An agentic skills framework & software development methodology that works.

234,96620,863Shell
AI 智能体ai-agentsbrainstorming
查看详情

hermes-agent

by NousResearch

10

The agent that grows with you

234,43747,175Python
AI 智能体ai-agentsagent-orchestration
查看详情

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

185,94028,768JavaScript
AI 智能体ai-agentsanthropicclaude-code
查看详情

cc-switch

by farion1231

3

A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Grok Build & Hermes Agent. Only official website: ccswitch.io

128,8688,826Rust
AI 智能体claude-codeai-tools
查看详情

claude-code

by anthropics

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.

120,03119,897Shell
AI 智能体
查看详情

开发者还喜欢

基于喜欢此 Skill 的开发者投票和收藏

ECC

by affaan-m

10

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

242,21936,702JavaScript
AI 智能体ai-agentsanthropicclaude-code
查看详情
15

An agentic skills framework & software development methodology that works.

234,96620,863Shell
AI 智能体ai-agentsbrainstorming
查看详情

hermes-agent

by NousResearch

10

The agent that grows with you

234,43747,175Python
AI 智能体ai-agentsagent-orchestration
查看详情

n8n

by n8n-io

12

Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.

201,88160,308TypeScript
MCP 服务器apisai-tools
查看详情

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

185,94028,768JavaScript
AI 智能体ai-agentsanthropicclaude-code
查看详情

cc-switch

by farion1231

3

A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Grok Build & Hermes Agent. Only official website: ccswitch.io

128,8688,826Rust
AI 智能体claude-codeai-tools
查看详情