BTCvision MCP Server

Model Context Protocol (MCP) server that exposes Bitcoin analytics, AI price predictions, halving countdowns, market signals, and crypto news from btcvision.org as tools for AI assistants.

Features

The server exposes 10 tools that any MCP-compatible client (Claude Desktop, Claude Code, Cursor, Cline, Continue, etc.) can call:

Installation

1. Clone & install

git clone https://github.com/your-user/btcvision-mcp.git
cd btcvision-mcp
python -m venv .venv
source .venv/bin/activate
pip install -e ".[dev]"

2. Run standalone (for debugging)

btcvision-mcp
# or
python -m btcvision_mcp.server

The server communicates over stdio (the standard MCP transport), so it will appear to "hang" — that's expected. It waits for JSON-RPC messages from an MCP client.

Configuration

Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "btcvision": {
      "command": "python",
      "args": ["-m", "btcvision_mcp.server"],
      "cwd": "/absolute/path/to/btcvision-mcp"
    }
  }
}

If you installed the package in a virtual environment, point command at the venv's Python:

{
  "mcpServers": {
    "btcvision": {
      "command": "/absolute/path/to/btcvision-mcp/.venv/bin/python",
      "args": ["-m", "btcvision_mcp.server"]
    }
  }
}

Claude Code

In .mcp.json at your project root:

{
  "mcpServers": {
    "btcvision": {
      "command": "python",
      "args": ["-m", "btcvision_mcp.server"]
    }
  }
}

Cursor / Cline / Continue

The same JSON shape works in their respective MCP config files.

Usage Examples

Once configured, you can ask your AI assistant things like:

• "What's Bitcoin's price right now according to btcvision?"
• "Show me the BTC price predictions for 2027 through 2030."
• "How many days until the next Bitcoin halving?"
• "What bullish signals is the BTCvision AI tracking?"
• "What's the current Fear & Greed Index?"
• "What's the BTCvision AI model version and accuracy?"
• "Look up: What is the Bitcoin halving?"
• "Give me the latest crypto news."

The assistant will call the relevant tool and return structured data.

Architecture

btcvision-mcp/
├── btcvision_mcp/
│   ├── __init__.py
│   ├── models.py       # Pydantic data models
│   ├── scraper.py      # Async HTTP scraping + parsing
│   └── server.py       # MCP server (stdio transport)
├── tests/
│   ├── __init__.py
│   └── test_server.py  # Unit + integration tests
├── pyproject.toml
├── README.md
└── .gitignore

• scraper.py — async web scraper with a 5-minute in-memory cache. Falls back to hard-coded signals and FAQ content if the site structure changes.
• models.py — strongly-typed Pydantic models for every response.
• server.py — the MCP server, using mcp.server.Server with stdio transport. 10 tools, each with an explicit JSON schema.
• tests/ — 25+ tests covering scraping, parsing, tool dispatch, and an end-to-end in-process MCP session.

Development

Run the tests:

pytest

Run the linter:

ruff check .

Type-check:

mypy btcvision_mcp

Tool Return Format

All tools return JSON-formatted text so any MCP client can render the output directly. Example:

{
  "ai_confidence_percent": 82.0,
  "projected_roi_multiplier": 3.4,
  "long_term_target": { "year": 2030, "price_usd": 350000.0 },
  "predictions": [
    { "year": 2027, "price_usd": 130000.0, "percent_change": 26.0,
      "drivers": "Post-2024-halving supply reduction, ETF inflows" },
    { "year": 2028, "price_usd": 200000.0, "percent_change": 94.0,
      "drivers": "4th Halving supply shock (block reward cut to 1.5625 BTC)" },
    { "year": 2029, "price_usd": 280000.0, "percent_change": 172.0,
      "drivers": "Bull run peak, new ATH" },
    { "year": 2030, "price_usd": 350000.0, "percent_change": 240.0,
      "drivers": "Post-cycle consolidation, nation-state adoption" }
  ],
  "source": "btcvision.org"
}

Caching

The scraper caches the homepage HTML for 5 minutes per process to avoid hammering the site. Use force=True (or restart the server) to bust the cache.

License

MIT — see LICENSE.

Disclaimer

All data is fetched from btcvision.org and is provided for educational and informational purposes only. It is not financial advice. Cryptocurrency investments carry significant risk. Always do your own research.