Skip to content
MCP server

VerifiMind PEAS - RefleXion Trinity

Three-judge LLM critique of a concept for innovation, ethics and security via MCP

Install
Pricing
Open source
Platforms
API
Runs
Remotely
Version
3.45.0
Listing updated
today

Similar MCP servers

About VerifiMind PEAS - RefleXion Trinity

VerifiMind PEAS - RefleXion Trinity is an open-source MCP server for ai models and llm tooling. It runs as a hosted remote server, so there is nothing to install locally.

VerifiMind PEAS runs an idea through three specialised judges in sequence. X looks at innovation and competitive positioning, Z covers ethics and compliance across a 21-framework jurisdictional check (GDPR and the EU AI Act among them), and CS reviews security against the OWASP Agentic AI Top 10. Each later stage receives the structured analysis of the one before, and the result is a unified assessment with scores, findings and a PROCEED, REVISE or REJECT recommendation.

The README is careful about what this is: structured multi-LLM critique, not formal verification or proof. It supports several model vendors, including Gemini, Claude, GPT, Groq, Cerebras, Mistral and Ollama. If a provider fails, only that stage degrades and incomplete runs withhold aggregate scores.

Of 13 tools, 8 are active; coordination handoffs and custom-template registration are disabled after security incidents. A free-forever pledge covers the tools, with rate limits of 10 requests per 60 seconds anonymously. Accounts are in maintenance, though the README says the four execution tools will later need a free registration.

Install

Run in a terminal

claude mcp add --transport http verifimind-genesis https://verifimind.ysenseai.org/mcp/

Add to Settings → Connectors → Add custom connector

https://verifimind.ysenseai.org/mcp/
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "verifimind-genesis": {
      "url": "https://verifimind.ysenseai.org/mcp/"
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "verifimind-genesis": {
      "type": "http",
      "url": "https://verifimind.ysenseai.org/mcp/"
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "verifimind-genesis": {
      "serverUrl": "https://verifimind.ysenseai.org/mcp/"
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "verifimind-genesis": {
      "type": "http",
      "url": "https://verifimind.ysenseai.org/mcp/"
    }
  }
}

Using several servers? Build one config file with the MCP config generator.

Tools this server provides (8)

Taken from the project documentation; the running server may differ.

consult_agent_x
Innovation analysis with competitive positioning.
consult_agent_z
Ethics review with 21-framework jurisdictional coverage.
consult_agent_cs
Security validation against the OWASP Agentic AI Top 10.
run_full_trinity
Runs X, Z and CS in sequence with a unified assessment.
list_prompt_templates
Browses templates by agent, category or tag.
get_prompt_template
Retrieves a template by ID.
export_prompt_template
Exports a template to Markdown or JSON.
get_template_statistics
Returns registry statistics by agent, phase and type.

What it does

Three specialised judges
Innovation, ethics and security agents review a concept in order.
Jurisdictional ethics check
Z agent covers 21 frameworks such as GDPR and the EU AI Act.
Security review
CS agent validates against the OWASP Agentic AI Top 10.
Staged context
Later stages receive the prior stage's structured analysis.
Prompt templates
Browse, retrieve and export built-in templates.
Multi-vendor models
Works across several LLM providers, with graceful degradation on failure.

Good for

  • Stress-test a product concept before building it
  • Get an ethics and compliance first pass across jurisdictions
  • Run a security review framed on agentic risks
  • Export a prompt template to Markdown or JSON

Pricing

Open source

Free, open source; anonymous limit 10 requests per 60 seconds

VerifiMind PEAS - RefleXion Trinity FAQ

Is it free?

Yes. The README pledges the validation tools stay free; rate limits exist only for system health.

Is it formal verification?

No. The output is structured multi-LLM critique, not a mathematical proof.

Which tools are available right now?

8 of 13 are active: the four Trinity tools and four template tools; coordination and custom-template mutation are disabled.

What is the connection URL?

https://verifimind.ysenseai.org/mcp/ with streamable-http and the trailing slash.

Is VerifiMind PEAS - RefleXion Trinity free?

Yes. VerifiMind PEAS - RefleXion Trinity is open source, so you can use and self-host it for free.

How do I add VerifiMind PEAS - RefleXion Trinity to Claude Code?

Run this command in your terminal: claude mcp add --transport http verifimind-genesis https://verifimind.ysenseai.org/mcp/

What are the best alternatives to VerifiMind PEAS - RefleXion Trinity?

The closest alternatives to VerifiMind PEAS - RefleXion Trinity are REA, headroom, omniget, next-ai-draw-io and mcp-for-blender.

Where this comes from

  • Description: written from the project's README.
  • Install steps: from the official MCP Registry entry.

How AIPoweredBot checks servers