NCCN Monitor

Monitor NCCN clinical guideline updates and get notified when new versions are published.

NCCN (National Comprehensive Cancer Network) publishes clinical practice guidelines for 92 cancer types and supportive care topics. These guidelines are updated frequently, and keeping track of changes is important for oncologists, researchers, and pharmaceutical professionals.

NCCN Monitor tracks version changes across all 92 NCCN professional guidelines, downloads updated PDFs, extracts update notes, and generates AI-powered change summaries — all accessible through the MCP protocol for integration with OpenClaw or any MCP-compatible AI assistant.

Features

• Complete coverage — monitors all 92 NCCN professional guidelines across 4 categories (Cancer by Type, Screening, Supportive Care, Specific Populations)
• Version change detection — compares current versions against stored state, detects updates
• PDF download + analysis — logs into NCCN (free registration), downloads updated PDFs, extracts update notes from the first few pages
• AI-powered summaries — generates structured prompts for LLM-based change analysis (Chinese or English)
• Health monitoring — detects silent scraping failures and alerts after consecutive errors
• Chinese language support — all 92 guidelines have Chinese names and aliases for fuzzy search
• Watch list management — configure which guidelines to monitor via natural language (Chinese or English)
• On-demand PDF download — download any guideline PDF by name (Chinese or English fuzzy search)
• Version archiving — automatically archive old versions when updates are detected, with standardized filenames (NCCN_GastricCancer_2026.V2_EN.pdf)
• Change timeline — view the full version history for any guideline

Quick Start

Prerequisites

• Python 3.11+
• uv (recommended) or pip
• NCCN account (free registration at nccn.org)

Install

git clone https://github.com/fshaan/nccn-monitor.git
cd nccn-monitor
uv venv --python 3.13 .venv
source .venv/bin/activate
uv pip install -e ".[dev]"

Configure

cp config.example.yaml config.yaml

Edit config.yaml to set your NCCN credentials and watch list. Or use environment variables:

export NCCN_USERNAME="[email protected]"
export NCCN_PASSWORD="your_password"

Run

# Start the MCP server (stdio transport)
python -m nccn_monitor.server

# Run unit tests
python -m pytest tests/ -v

# Run integration test (needs NCCN credentials)
python tests/integration_test.py

MCP Tools

OpenClaw Integration

Native Plugin (recommended)

Install as an OpenClaw plugin for direct tool registration:

# Copy plugin to OpenClaw extensions
cp -r openclaw-plugin ~/.openclaw/extensions/nccn-monitor
cd ~/.openclaw/extensions/nccn-monitor && npm install

Then add to your ~/.openclaw/openclaw.json under plugins.entries:

{
  "nccn-monitor": {
    "enabled": true,
    "config": {
      "pythonPath": "/path/to/nccn-monitor/.venv/bin/python",
      "projectDir": "/path/to/nccn-monitor"
    }
  }
}

After openclaw gateway restart, 8 tools (nccn_check_updates, nccn_get_status, etc.) appear natively in OpenClaw.

Note: Only keys declared in openclaw.plugin.json configSchema are allowed in the plugin config. Settings like archive_dir go in the project's config.yaml, not in openclaw.json.

Via mcporter (alternative)

If you prefer using mcporter as middleware:

npx mcporter config add nccn-monitor --transport stdio \
  --command "/path/to/.venv/bin/python" --args "-m" --args "nccn_monitor.server"

Then use the mcporter skill in OpenClaw to call tools.

Cron scheduling

Set up a daily cron job in OpenClaw to call nccn_check_updates (or check_updates via mcporter) for automatic monitoring.

Architecture

OpenClaw (cron + notifications)
    │
    │ MCP Protocol (stdio)
    ▼
┌─────────────────────────────────────┐
│      NCCN Monitor MCP Server        │
│                                     │
│  scraper ──→ state ──→ notifier     │
│     │                               │
│     ▼ (on version change)           │
│  downloader ──→ analyzer            │
│  (NCCN login)   (PDF extraction)    │
└─────────────────────────────────────┘

Credits

• NCCN login and PDF download logic adapted from gscfwid/NCCN_guidelines_MCP
• Built with FastMCP for MCP server implementation

License

MIT