菜单
The Claw Fleet — 5 AI engines (Media, Forge, Herald, Scout, Scribe) with 33 skills, cross-claw events, and fleet orchestration for creative operations.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Visual web dashboard for managing AI agent pipelines. Inbox with approval cards, live agent monitoring, social scheduling, and publish pipeline.
Complete AI-powered media processing pipeline for creative world-building. Scans, classifies, processes, scores, and uploads media with fantasy/mythology-aware tagging.
Auto-generate platform-optimized image variants and AI-written captions for Instagram, LinkedIn, X, YouTube, and TikTok from any source image.
AI-powered aesthetic quality scoring for images. 5-dimension evaluation: Canon Alignment, Design Compliance, Emotional Impact, Technical Fit, Uniqueness. Scores 0-100 with automatic tier assignment.
Classify untagged media against Arcanea canon using Gemini Vision
Detect and reject duplicate media files by content hash
| name | arcanea-claw |
| version | 0.3.0 |
| description | The Claw Fleet — 5 AI engines (Media, Forge, Herald, Scout, Scribe) with 33 skills, cross-claw events, and fleet orchestration for creative operations. |
| author | frankxai |
| homepage | https://arcanea.ai/claw |
| repository | https://github.com/frankxai/arcanea-claw |
| license | MIT |
| tags | ["media","image-processing","ai-classification","nft","social-media","creative","world-building","publishing","mcp","fleet"] |
| platforms | ["openclaw","nanoclaw","claude-code","standalone"] |
| install | ["pip install -r requirements.txt","cd mcp-server && npm install && npm run build"] |
| mcpServer | {"command":"node","args":["mcp-server/dist/index.js"],"env":{"CLAW_DIR":"${CLAW_DIR:-./}","MANIFEST_PATH":"${MANIFEST_PATH:-./manifest.json}"}} |
| metadata | {"openclaw":{"requires":{"env":["SUPABASE_URL","SUPABASE_SERVICE_KEY","GEMINI_API_KEY"],"bins":["python3","ffmpeg"],"anyBins":["docker","podman"]},"primaryEnv":"SUPABASE_URL","emoji":"lobster","homepage":"https://arcanea.ai/claw","security":{"sandboxed":true,"containerOnly":true,"permissions":["network:outbound","storage:supabase","storage:vercel-blob"],"noPermissions":["shell:execute","filesystem:write-outside-data"]},"verified":true,"publisher":"@arcanea"}} |
A complete AI-powered pipeline for scanning, classifying, scoring, processing, and publishing creative media. Built for world-builders, game designers, artists, and anyone managing large collections of visual content.
Source Files --> Scan --> Classify --> Score --> Process --> Upload
(any dir) (discover) (tag/label) (TASTE) (resize/opt) (cloud)
8 skills working as a pipeline:
| Skill | What it does |
|---|---|
media-scan | Discovers new images, video, and audio in configured directories |
media-classify | Tags assets by character, element, content type using heuristics + AI |
taste-score | Scores aesthetic quality 0-100 across 5 dimensions using Gemini Vision |
media-dedup | Finds and marks duplicate assets by perceptual hash |
media-process | Resizes to 6 variants (hero, gallery, social square/story/wide, thumbnail) |
media-upload | Deploys hero-tier to Vercel Blob, gallery-tier to Supabase Storage |
social-prep | Generates platform-optimized variants with AI captions and hashtags |
notify | Sends pipeline status updates via webhook |
openclaw install @arcanea/claw
This registers the MCP server and all 8 skills automatically via openclaw.json.
nanoclaw add arcanea-claw
Add to your .mcp.json:
{
"mcpServers": {
"arcanea-claw": {
"command": "node",
"args": ["/path/to/arcanea-claw/mcp-server/dist/index.js"],
"env": {
"CLAW_DIR": "/path/to/media",
"MANIFEST_PATH": "/path/to/manifest.json"
}
}
}
}
git clone https://github.com/frankxai/arcanea-claw.git
cd arcanea-claw
./install.sh
Once installed, these tools are available to any MCP-compatible agent:
| Tool | Description |
|---|---|
claw_status | Engine status: daemon, manifest, config |
claw_scan | Scan directories for new media files |
claw_inbox | List assets filtered by status/guardian with counts |
claw_approve | Approve assets by ID |
claw_reject | Reject assets by ID with optional reason |
claw_stats | Full dashboard: guardians, elements, tiers, scores |
claw_process | Process approved assets (resize, convert) |
claw_heroes | List hero-tier assets (score 80+) |
claw_classify | Classify untagged assets by filename heuristics |
claw_pipeline | Run full pipeline: scan -> classify -> dedup -> process |
Edit config.yaml to customize:
arcanea_claw:
scan:
paths:
- /your/media/directory
extensions: [.jpg, .jpeg, .png, .webp, .gif, .mp4]
process:
webp_quality: 85
max_dimension: 2400
size_variants:
hero: [1920, 1080]
gallery: [1200, 800]
social_square: [1080, 1080]
social_story: [1080, 1920]
social_wide: [1200, 628]
thumbnail: [320, 320]
upload:
hero_threshold: 80 # TASTE score for Vercel Blob
gallery_threshold: 60 # TASTE score for Supabase Storage
classify:
model: "gemini-2.0-flash"
| Variable | Required | Description |
|---|---|---|
CLAW_DIR | No | Root directory for media operations (default: cwd) |
MANIFEST_PATH | No | Path to manifest.json (default: $CLAW_DIR/manifest.json) |
GEMINI_API_KEY | For scoring | Google Gemini API key for TASTE scoring and classification |
SUPABASE_URL | For upload | Supabase project URL |
SUPABASE_SERVICE_ROLE_KEY | For upload | Supabase service role key |
BLOB_READ_WRITE_TOKEN | For upload | Vercel Blob read/write token |
NOTIFY_WEBHOOK_URL | For notify | Webhook URL for pipeline notifications |
arcanea-claw/
mcp-server/ # MCP tools (Node.js, TypeScript)
index.ts # 11 tools, stdio transport
dist/ # Compiled output
engine/ # Python pipeline engine
daemon.py # Background pipeline runner
skills/ # Python skill implementations
skills/ # Skill definitions (8 skills)
skillpacks/ # Bundled skill collections
config.yaml # Pipeline configuration
openclaw.json # OpenClaw integration config
clawhub.json # ClawHub marketplace listing
install.sh # Universal installer
MIT