clawned（技能安全扫描）

Clawned — Security Agent for OpenClaw

Automatically discovers all installed skills, analyzes them for security threats, and syncs results to your Clawned dashboard.

Setup

Configure your API key in openclaw.json:

{
  "skills": {
    "entries": {
      "clawned": {
        "enabled": true,
        "env": {
          "CLAWNED_API_KEY": "cg_your_api_key_here",
          "CLAWNED_SERVER": "https://api.clawned.io"
        }
      }
    }
  }
}

Commands

Sync all installed skills to dashboard

python3 {baseDir}/scripts/agent.py sync

Scan a single skill locally

python3 {baseDir}/scripts/agent.py scan --path <skill-directory>

List all discovered skills

python3 {baseDir}/scripts/agent.py inventory

Check agent status

python3 {baseDir}/scripts/agent.py status

Data & Privacy

During sync (default operation):

• Sends only skill metadata: name, owner, slug, version
• No file contents are uploaded
• No .env files or secrets are ever read

During scan --path (explicit user action only):

• Reads source files (.md, .py, .js, etc.) from the specified skill directory for analysis
• .env files are excluded from scanning
• File contents are sent to the Clawned server for security analysis

Local config access:

• Reads ~/.openclaw/openclaw.json only to locate skill directories (extraDirs paths)
• No credentials or secrets are read from config files

Automatic Sync

Schedule every 6 hours via OpenClaw cron:

{
  "jobs": [
    {
      "schedule": "0 */6 * * *",
      "command": "Run clawned sync to check all installed skills",
      "description": "Security scan every 6 hours"
    }
  ]
}