GitHub Events Monitor

A Python-based service that monitors GitHub Events (WatchEvent, PullRequestEvent, IssuesEvent), stores them in SQLite, exposes data through an MCP server (each tool = API endpoint), and publishes interactive visualizations to a GitHub Pages site from the docs/ folder.

Architecture (C4 L1)

[Person: Dashboard User] → [Container: GitHub Pages (docs/)] ← [Container: Data Exporter] ← [Container: Events DB (SQLite)] ← [Container: Event Monitor Service] ← [External System: GitHub API /events] [Person: API Consumer] → [Container: MCP Server] → [Container: Events DB (SQLite)]

Components

• Event Monitor: Polls https://api.github.com/events, filters target types, stores to SQLite, computes PR metrics.
• MCP Server: Tools to retrieve metrics/data directly from the DB.
• Data Exporter: Builds docs/data.json and Plotly HTML charts in docs/.
• GitHub Pages: Static dashboard served from docs/.

Setup

1. Python env

• python -m venv venv
• source venv/bin/activate
• pip install -r requirements.txt

2. Initialize DB

• sqlite3 database/events.db < database/schema.sql

3. Run monitor

• python service/github_monitor.py # continuous
• python service/github_monitor.py --once # single cycle (useful in CI) Optional: export GITHUB_TOKEN for higher limits.

4. Export dashboard data

• python service/data_exporter.py

5. MCP server

• python mcp/server.py

GitHub Pages

• Push to main branch.
• Enable Pages from Settings → Pages → Deploy from a branch → main → /docs.

Notes

• created_at_ts (epoch) enables correct time filtering and SQLite date ops.
• For production, consider PostgreSQL, containerization, and a retention policy for events.