MyXstack

This repository hosts a lightweight, step-by-step guide for setting up an autonomous X (Twitter) agent system that acts based on thread context & reasoning, through Grok via the xMCP server.

Phase 1: Gather prerequisites & accounts (1–2 hours)

X developer account & app

1. Go to https://developer.x.com and sign in with the account you want the bot to run on.
2. Create a new app/project (free tier is fine to start).
3. Set app permissions to Read + Write + Direct Messages so the agent can post replies.
4. Generate and record the OAuth 1.0a credentials:
   • API Key
   • API Secret
   • Access Token (user context)
   • Access Token Secret

Dedicated bot account (recommended)

1. Create a new X account for the bot (for example, @MyXMCPBot).
2. Link the bot account to your developer app via the OAuth flow.
3. Note the bot user ID (you can fetch this via the API later).

xAI / Grok API key

1. Visit https://console.x.ai, open the API keys section, and create a key that starts with xai-.
2. Store the key securely.

Local tooling

• Install Python 3.9+ (3.10–3.13 recommended).
• Ensure git is installed for cloning.

Phase 2: Clone & set up the xMCP server (local first)

1. Clone your fork of the xMCP server:

   git clone https://github.com/groupthinking/xMCP.git
   cd xMCP

2. Create and activate a virtual environment:

   python -m venv .venv
   source .venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

4. Create your .env:

   cp env.example .env

5. Fill in the .env values you need:

   # OAuth user token for posting replies
   X_OAUTH_ACCESS_TOKEN=your_access_token_here
   X_OAUTH_ACCESS_TOKEN_SECRET=your_access_token_secret_here
   
   # Optional read-only bearer token
   # X_BEARER_TOKEN=your_bearer_token_here
   
   # Grok / xAI
   XAI_API_KEY=your_xai_key_here
   XAI_MODEL=grok-4-1-fast
   
   # Optional tuning
   MCP_HOST=127.0.0.1
   MCP_PORT=8000
   X_API_DEBUG=1
   

6. If you need to generate OAuth user tokens, add CLIENT_ID and CLIENT_SECRET to .env, update the redirect URI in generate_authtoken.py, and run:

   python generate_authtoken.py

Phase 3: Test the core xMCP server + Grok connection

1. Start the MCP server:

   python server.py

   You should see Server running on http://127.0.0.1:8000.
2. In another terminal, test Grok integration:

   python test_grok_mcp.py

3. Try prompts like:
   • Search recent posts about MCP
   • Get my user info
   • Post a test reply saying 'Hello from MCP' to tweet ID 1234567890

If these work, the core pipe (Grok → xMCP → X API) is live.

Phase 4: Add the trigger layer (tag/mention → action)

A simple polling listener is provided in listener.py at the repo root. It uses Tweepy to poll for mentions and posts a placeholder reply. The listener requires OAuth user credentials (plus a bearer token for reads). Install Tweepy, export your environment variables, and run it while the xMCP server is up:

pip install tweepy
export X_API_KEY=your_api_key
export X_API_SECRET=your_api_secret
export X_OAUTH_ACCESS_TOKEN=your_access_token
export X_OAUTH_ACCESS_TOKEN_SECRET=your_access_token_secret
export X_BEARER_TOKEN=your_bearer_token
export POLL_INTERVAL_SECONDS=60
python listener.py

Then tag your bot from another account to see it respond. Replace the placeholder Grok call with your real xMCP/Grok invocation when ready.

Phase 5: Iterate toward full autonomy

• Wire the listener to call Grok through the xMCP server.
• Add Docker/Docker Compose if you want a containerized workflow.
• Persist state (SQLite) to avoid reprocessing old mentions.
• Extend with proxies, CrewAI/AutoGen, or other agent frameworks.
• Deploy to Railway/Render/VPS for 24/7 operation.

Current status

Start with Phases 1–4 today. If you hit a snag, share the exact output/error and we can debug together.