Sassy Fact Check

Dual MCP setup for Instagram DM bot that auto-adjusts tone and provides sourced fact-check responses.

python

GitHub Stars

python

Language

5 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "sangreal-007-sassy-factcheck-bot": {
      "command": "/Users/yourusername/.local/bin/uv",
      "args": [
        "run",
        "--directory",
        "/Users/yourusername/instagram_dm_mcp",
        "python",
        "src/mcp_server.py"
      ],
      "env": {
        "ENABLE_SAFE_MODE": "true",
        "LOG_INTERACTIONS": "true",
        "ANTHROPIC_API_KEY": "YOUR_CLAUDE_KEY",
        "INSTAGRAM_PASSWORD": "your_password",
        "INSTAGRAM_USERNAME": "your_username",
        "INSTAGRAM_REAL_MODE": "false"
      }
    }
  }
}

You operate a dual MCP server that powers an Instagram DM bot to fact-check health claims with cited sources and a distinct, sassy tone. It coordinates between an Instagram DM MCP and a Sassy Bot MCP to deliver timely, sourced responses while auto-adjusting tone for sensitive topics and logging interactions for analytics.

How to use

To use this MCP setup, run both servers so they can communicate in real time. The Instagram DM MCP handles conversations with real Instagram users, while the Sassy Bot MCP generates pointed, sourced responses. The flow is: you receive a DM in the Instagram MCP, the Sassy Bot MCP creates a sassy response with sources, and the Instagram MCP sends that reply back to the user.

Your configuration enables two distinct servers to run side by side. Real mode connects to Instagram via Gala Labs for live conversations, while demo mode uses practice claims to test the tone, flow, and citation generation without touching real accounts.

How to install

Prerequisites you need before installing the MCP setup:

Clone the project repository to your local machine.
Create and activate a Python virtual environment.
Install dependencies from the requirements file.
Prepare the dual MCP configuration in Claude Desktop to reference both servers.
Start both servers and verify the dual workflow by testing with demo data first, then switch to real mode if you have access to live Instagram.

Configuration and running tips

The setup uses two MCP entries in Claude Desktop, both configured to run as standard input/output (stdio) servers. Each server runs a Python script that powers its respective MCP.

Environment and mode controls are defined separately to switch between demo and real mode without changing the code paths. For testing, begin in demo mode to validate behavior before enabling real mode.

Security and usage notes

For live testing, use a dedicated test Instagram account to prevent automation from affecting your personal profile. Automation on Instagram can trigger restrictions, so keep testing separate from your main account.

The system is designed to tailor response tone automatically based on content sensitivity, while ensuring health misinformation is clearly debunked with citations.

Troubleshooting

If one server stops responding, restart both servers and verify Claude Desktop has the correct project paths loaded for the two MCPs. Confirm that the virtual environment is active for the Sassy Bot MCP and that the Instagram MCP has network access if you are testing live mode.

Examples of usage

# Start the two MCP servers from separate terminals per the guide
# Terminal 1: Instagram DM MCP is managed by Claude Desktop via the Instagram path
# Terminal 2: Sassy Bot MCP runs in the project environment
python src/mcp_server.py

Available tools

generate_sassy_response

Create viral fact-check responses with sources tailored to the detected content, including tone adjustments.

generate_welcome_message

Produce welcoming messages for new followers or first-time interactions.

check_instagram_dms

Show practice claims in demo mode or guide users toward the real Instagram MCP flow.

instagram_integration_status

Display the current status of the dual MCP integration and readiness.

list_chats

Fetch active conversations from the Instagram MCP (demo mode) or prepare for real-mode conversations.

send_message

Send generated responses back to Instagram users via the Instagram MCP.

get_user_followers

Inspect follower lists via the Instagram MCP for context on conversations.

get_thread_details

Retrieve full conversation histories from Instagram threads.

Built by

VeilStrat

AI signals for GTM teams