Proofly MCP Integration
The Proofly MCP server enables deepfake detection and face analysis of images.
You can:
Analyze images from a URL for deepfake detection
Analyze base64-encoded images for deepfake detection
Check the status of an analysis session using its UUID
Retrieve detailed information about specific faces detected in an analysis session
Provides a package (proofly-mcp) that acts as an MCP server for deepfake detection, installable via npm either globally or locally.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Proofly MCP Integration analyze this image for deepfakes: https://example.com/suspicious-photo.jpg"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Proofly MCP Integration
Install and just write 'proofly it' URL to content or analyze it URL to content for deepfake face swap analysis.
For clients that connect to MCP servers using a URL (e.g., Cursor, Cascade/Windsurf)
Add one of the following configurations to your MCP client (e.g., in mcp_config.json):
A. Streaming (SSE - Recommended where supported):
{
"proofly": {
"serverUrl": "https://mcp.proofly.ai/sse",
"supportedMethods": [
"analyze-image",
"analyze",
"get-face-details",
"check-session-status"
],
"auth": { "type": "none" } // Or your specific auth if Proofly API https:/get.proofly.ai requires it
}
}B. Standard HTTP (Non-streaming):
{
"proofly": {
"serverUrl": "https://mcp.proofly.ai/mcp",
"supportedMethods": [
"analyze-image",
"analyze",
"get-face-details",
"check-session-status"
],
"auth": { "type": "none" } // Or your specific auth if Proofly API https:/get.proofly.ai requires it
}
}For clients that can execute a local command for an MCP server (e.g., Claude Desktop)
Claude Desktop:
Run: npx proofly-mcp@latest
Add to your Claude Desktop config file (e.g.,
claude_desktop_config.json)
{
"mcpServers": {
"proofly": {
"command": "npx",
"args": [
"-y", // The -y flag might be specific to your npm/npx version or aliasing for auto-confirmation.
"proofly-mcp@latest"
],
"supportedMethods": [
"analyze-image",
"analyze",
"get-face-details",
"check-session-status"
]
}
}
}Alternatively, if you have proofly-mcp installed globally (npm install -g proofly-mcp), you can use:
{
"mcpServers": {
"proofly": {
"command": "proofly-mcp",
"args": [],
"supportedMethods": [
"analyze-image",
"analyze",
"get-face-details",
"check-session-status"
]
}
}
}Other command-capable MCP Clients:
If your MCP client can launch a local command, configure it to run proofly-mcp.
Conceptual example (actual config varies by client):
{
"mcpServers": {
"proofly": {
"type": "command",
"command": "proofly-mcp",
"supportedMethods": [
"analyze-image",
"analyze",
"get-face-details",
"check-session-status"
]
}
}
}Environment Variables for proofly-mcp CLI (Optional)
PROOFLY_API_KEY: Your Proofly API key. Theproofly-mcpCLI will use this API key if the variable is set when communicating with Proofly APIhttps://get.proofly.ai.
Available MCP Methods
analyze
Analyzes an image from a URL for deepfake detection.
analyze-image
Analyzes an image provided as a base64 string for deepfake detection.
check-session-status
Checks the status of a deepfake analysis session.
get-face-details
Gets detailed information about a specific face detected in an image analysis session.
Available Tools
4 toolsanalyzeB
Analyzes an image from a URL for deepfake detection.
Inputs: imageUrl (required), format.
analyze-imageC
Analyzes an image provided as a base64 string for deepfake detection.
Inputs: imageBase64 (required), filename (required), format.
check-session-statusC
Check the status of a deepfake analysis session.
Inputs: sessionUuid (required), format.
get-face-detailsC
Get detailed information about a specific face detected in an image.
Inputs: sessionUuid (required), faceIndex (required), format.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 4 tools
The tools 'analyze' and 'analyze-image' have significant overlap, as both perform deepfake detection on images but differ only in input format (URL vs. base64 string). This creates ambiguity, as an agent might misselect between them without clear guidance on which to use based on context. The other tools ('check-session-status' and 'get-face-details') are more distinct but the core analysis duplication undermines clarity.
The naming is mixed with some consistency issues: 'analyze' and 'analyze-image' use a verb-object pattern but differ in hyphenation, while 'check-session-status' and 'get-face-details' follow a verb-noun-noun pattern with hyphens. This creates a readable but inconsistent convention, as tools alternate between simple verbs and more descriptive phrases without a uniform style.
With 4 tools, the count is reasonable for a deepfake detection server, covering analysis, status checking, and detailed face information. It's slightly lean but not overly thin, as each tool serves a purpose without obvious bloat. A few more tools might enhance completeness, but the current scope is adequate for basic operations.
The toolset covers core analysis and status checking but has notable gaps: there are no tools for managing sessions (e.g., create or delete), handling batch processing, or providing summary reports. While agents can perform basic detection and retrieve details, the lack of lifecycle management tools limits workflow efficiency and may require workarounds.
Maintenance
Related MCP Connectors
An MCP Server that provides identity verification and anti-fraud tools for AI agents via deepidv.
MCP server for Flux AI image generation
MCP server for Wan AI video generation
MCP server for Qwen Image 3 AI image generation
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceAn MCP server for analyzing images using OpenRouter vision models, offering capabilities like automatic image resizing, model configuration, and handling custom queries about images.11MIT
- AlicenseNot gradedqualityCmaintenanceDeepfake Detector - MCP server providing AI-powered tools and automation by MEOK AI Labs10 npm44 PyPIMIT
- AlicenseAqualityDmaintenanceAn MCP server that enables AI assistants to analyze images for AI-generated content using noise maps, error level analysis, frequency analysis, spectral decay, color analysis, and metadata inspection.7MIT
- AlicenseAqualityDmaintenanceMCP server providing image analysis tools for AI agents, including metadata extraction, favicon discovery, and placeholder generation.557 npmMIT