Resemble Detect Skill

An agent skill for deepfake detection and media safety — powered by Resemble AI.

Give any AI agent the ability to detect AI-generated audio, images, and video using Resemble's detection platform.

Compatible Agents

Works with any agent that supports markdown skills:

Agent	Install Method
Claude Code	npx skills add resemble-ai/detect-skill or copy to .claude/skills/
OpenClaw (formerly Clawdbot)	Copy to skills directory or import via built-in skill loader
Hermes Agent	Add to skills directory — Hermes will auto-discover and self-improve on it
Cursor	Add to .cursor/skills/ or project rules
GitHub Copilot	Add to .github/copilot-instructions.md or reference in prompt
Windsurf	Add to project rules
Gemini CLI	Add to .gemini/skills/

Install

Via skills.sh (recommended):

npx skills add resemble-ai/detect-skill

Manual: Copy SKILL.md into your agent's skills directory.

What It Does

This skill teaches your agent to use Resemble's full detection and media safety suite:

Capability	Description
Deepfake Detection	Analyze audio, image, and video for synthetic manipulation with confidence scores and visualizations
Intelligence	Extract speaker info, emotion, transcription, misinformation signals, and abnormalities from any media
Detect Intelligence	Ask natural-language follow-up questions about completed detection results
Audio Source Tracing	Identify which AI platform (ElevenLabs, Resemble, etc.) synthesized detected fake audio
Watermarking	Apply invisible watermarks for provenance tracking, or detect existing watermarks in media
Identity Verification	Create voice profiles and match unknown speakers against them (Beta)

Requirements

• A Resemble AI API key
• Media files accessible via public HTTPS URLs

Pair With the Resemble MCP Server (Recommended)

For the richest experience, pair this skill with the Resemble MCP server so your agent gets live access to Resemble's full documentation, API reference, and endpoint schemas — no guessing, no stale examples.

Easiest setup — hosted SSE endpoint (zero install)

https://mcp.resemble.ai/sse

Point any MCP-compatible agent at this URL and you're done. No cloning, no installing, no server to run.

Cursor — add to .cursor/mcp.json:

{
  "mcpServers": {
    "resemble": {
      "url": "https://mcp.resemble.ai/sse"
    }
  }
}

Claude Desktop / Claude Code — add to claude_desktop_config.json (or .claude/mcp.json):

{
  "mcpServers": {
    "resemble": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://mcp.resemble.ai/sse"]
    }
  }
}

Any other MCP-compatible agent (OpenClaw, Hermes Agent, Windsurf, Cline, Continue, etc.) — point the MCP client at https://mcp.resemble.ai/sse. See the Resemble MCP README for per-agent config snippets.

Prefer self-hosting? Clone resemble-ai/resemble-mcp and run it locally.

Once connected, the agent gets tools like resemble_docs_lookup, resemble_api_endpoint, and resemble_api_search — this skill is built to use them when available.

How It Works

The skill is a single markdown file (SKILL.md) that provides your AI agent with:

• Decision tree — maps user intent to the correct API capability and endpoint
• Complete API reference — every endpoint, parameter, and response schema
• Score interpretation — how to read and present detection confidence scores
• Workflow templates — full forensics, quick checks, and provenance pipelines
• Red flags — anti-patterns the agent should catch in its own reasoning
• Error handling — every status code with cause and resolution

Agents like Hermes Agent with self-improving skill systems will automatically refine their use of this skill over time. OpenClaw's 100+ prebuilt skills ecosystem makes it a natural fit — drop detect.md in and it works alongside existing skills immediately.

Example Prompts

Once installed, try asking your agent:

• "Is this audio file a deepfake?"
• "Analyze this video for AI manipulation and tell me what platform might have generated it"
• "Apply a watermark to this image for provenance tracking"
• "Check if this speaker matches the voice profile we have on file"
• "What can you tell me about this audio — speaker, emotion, language, any abnormalities?"

Links

• Resemble AI — Platform
• mcp.resemble.ai/sse — Hosted MCP endpoint (zero install)
• resemble-ai/resemble-mcp — Self-host the MCP server
• API Documentation — Resemble API docs
• skills.sh — The Open Agent Skills Ecosystem
• OpenClaw — Open-source AI agent (formerly Clawdbot)
• Hermes Agent — Self-improving AI agent by Nous Research

License

Apache-2.0