ghstats

ghstats is a lightweight, self-hosted dashboard written in Rust that collects GitHub repository traffic metrics and stores historical data beyond GitHub's default 14-day retention. It runs as a single small container and persists metrics in a local SQLite database for per-repo dashboards and simple API access.

Key Features

• Periodic collection of GitHub traffic metrics (views, uniques, clones) and daily aggregation stored in SQLite
• Per-repository listing and simple dashboard with historical metrics
• Minimal single-container deployment with a small Docker image and optional Docker Compose configuration
• Environment-driven configuration: GitHub token, host/port, filters, custom header links, and API token
• Flexible repo filtering (include/exclude rules, wildcards, meta rules like !fork and !archived)
• Hourly synchronization scheduler and immediate initial sync when database is empty
• Simple JSON API endpoint (/api/repos) protected by an API token for programmatic access
• Option to include private repos when using a token with appropriate scopes

Use Cases

• Preserve and analyze GitHub repository traffic history beyond the platform's 14-day view window
• Build lightweight public or internal dashboards showing repository popularity and trends
• Export or integrate basic repo traffic metrics via the built-in JSON API for reporting or automation

Limitations and Considerations

• Dependent on GitHub API rate limits and token scopes; private repo collection requires broader token permissions
• Uses SQLite and a single-container design, which limits horizontal scalability for very large numbers of repositories
• Data granularity is daily; current-day data is updated by re-fetching but there is no finer time-series resolution
• API surface is minimal (primarily a consolidated /api/repos endpoint) compared to full analytics platforms

ghstats is suitable when you need a simple, low-dependency way to retain and inspect GitHub traffic history. It favors minimal operational complexity over advanced analytics or large-scale multi-node deployments.