Skip to content
MCP server

RogueGPT

Store and retrieve controlled AI news stimuli with provenance for misinformation research

Runs locallyStatus and history
Install
Pricing
Open source
Platforms
CLI, self-hosted
Runs
Locally
Version
1.5.0
Listing updated
3 days ago

About RogueGPT

RogueGPT is an open-source MCP server for ai models and llm tooling. It runs locally on your machine and is started by your AI client.

RogueGPT is research instrumentation, and this MCP server is its fragment store. Fragments are short, fully parameterised pieces of synthetic news that researchers use to study how people and detectors respond to fabricated articles. Each fragment records the model, prompt, language, style and format that produced it, and the server needs a MongoDB connection string supplied as ROGUEGPT_MONGO_URI.

The README describes a corpus of 3,278 multilingual fragments (2,638 machine-made, 640 human-sourced) from 10 models across 6 providers, in 4 languages, 3 formats and 5 journalistic styles per language. RogueGPT is the upstream stage of a pipeline: CRED-1 flags unreliable sources, RogueGPT produces stimuli, and JudgeGPT gathers human judgements tied to the generation parameters.

The maintainers state it has no publishing, posting, distribution, scheduling or audience-targeting capability. Researchers use their own API credentials, the corpus is archived under restricted academic access, and the software is GPL-3.0. It is described in the Journal of Open Source Software and installs with pip install roguegpt[mcp].

Install

Run in a terminal

claude mcp add roguegpt --env ROGUEGPT_MONGO_URI='<ROGUEGPT_MONGO_URI>' -- uvx roguegpt

Add to claude_desktop_config.json

{
  "mcpServers": {
    "roguegpt": {
      "command": "uvx",
      "args": [
        "roguegpt"
      ],
      "env": {
        "ROGUEGPT_MONGO_URI": "<ROGUEGPT_MONGO_URI>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "roguegpt": {
      "command": "uvx",
      "args": [
        "roguegpt"
      ],
      "env": {
        "ROGUEGPT_MONGO_URI": "<ROGUEGPT_MONGO_URI>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "roguegpt": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "roguegpt"
      ],
      "env": {
        "ROGUEGPT_MONGO_URI": "<ROGUEGPT_MONGO_URI>"
      }
    }
  }
}

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

{
  "mcpServers": {
    "roguegpt": {
      "command": "uvx",
      "args": [
        "roguegpt"
      ],
      "env": {
        "ROGUEGPT_MONGO_URI": "<ROGUEGPT_MONGO_URI>"
      }
    }
  }
}

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

{
  "mcpServers": {
    "roguegpt": {
      "command": "uvx",
      "args": [
        "roguegpt"
      ],
      "env": {
        "ROGUEGPT_MONGO_URI": "<ROGUEGPT_MONGO_URI>"
      }
    }
  }
}

Configuration

  • ROGUEGPT_MONGO_URIMongoDB connection string for the fragment storeRequired

What it does

Full provenance
Every fragment keeps its model, prompt, language, style and format.
Multi-model corpus
Fragments come from 10 models across 6 providers.
Multilingual and styled
Covers 4 languages, 3 formats and 5 journalistic styles per language.
Simulation seeding
Standardises seed content for agent-based misinformation diffusion studies.
Pipeline links
Ties stimuli to human-perception data collected in JudgeGPT.
Research safeguards
No publishing or targeting features; restricted academic corpus access.

Good for

  • Pull controlled news stimuli for a perception experiment
  • Seed an agent-based misinformation simulation reproducibly
  • Red-team a fake-news detector with parameterised variants
  • Trace which generation settings produced a given fragment

Pricing

Open source

GPL-3.0

Prices change. Check aloth.github.io before you buy.

RogueGPT FAQ

What does the server need to run?

A MongoDB connection string in ROGUEGPT_MONGO_URI, installed through pip install roguegpt[mcp].

Can it publish or distribute content?

No. The README says it has no publishing, posting, distribution, scheduling or audience-targeting capability.

What licence applies?

GPL-3.0 for the software, with the corpus archived under restricted academic access.

Who is it for?

Misinformation researchers, red-teamers, agent-based simulation researchers and human-factors researchers.

Is RogueGPT free?

Yes. RogueGPT is open source, so you can use and self-host it for free.

How do I add RogueGPT to Claude Code?

Run this command in your terminal: claude mcp add roguegpt --env ROGUEGPT_MONGO_URI='<ROGUEGPT_MONGO_URI>' -- uvx roguegpt

What are the best alternatives to RogueGPT?

The closest alternatives to RogueGPT are Agent-Native Analytics, Claude Code Ultimate Guide, Google Ads + Meta Ads + SEO/GEO MCP, Tapetide - Indian Stock Market MCP and World Monitor.

Alternatives to RogueGPT

  1. Open-source analytics app where an AI agent builds charts and dashboards from your data

    • Remote
    • Open source
    • 7.7K GitHub stars
    • Analytics and data
  2. MCP server for searching the Claude Code Ultimate Guide from any MCP client

    Runs locally
    • Open source
    • 6.5K GitHub stars
    • AI models and LLM tooling
  3. Hosted MCP connector for managing Google Ads, Meta Ads, GA4 and Search Console from chat

    • 8 tools
    • Remote
    • Freemium
    • 8.1K GitHub stars
    • Marketing and SEO
  4. Indian NSE and BSE stock market data for AI assistants through 48 MCP tools

    • 17 tools
    • Remote
    • Freemium
    • 150 GitHub stars
    • Finance and payments
  5. World Monitor

    MCP server

    Live geopolitical intelligence dashboard with an API and MCP server for agents

    • Remote
    • Freemium
    • Analytics and data
  6. CompanyData

    MCP server

    Search and enrich company records from official registers via the CompanyData API

    • 3 tools
    • Remote
    • Paid
    • 3 GitHub stars
    • Analytics and data