txtop documentation

Start with see your first usage data. It installs txtop, reads local coding-agent history without an API key, and introduces the dashboard. The project overview is the shorter install and feature summary.

Tutorial

• See your first usage data; go from installation to a readable local-usage dashboard.

How-to guides

• Connect a provider API; add OpenAI, Anthropic, or OpenRouter organization usage and cost.
• Track local coding-agent usage; scan Codex, Claude Code, and Gemini CLI history.
• Enable live subscription limits; opt into experimental credential-backed 5-hour and 7-day windows.
• Ingest OpenTelemetry; connect a coding agent or another OTLP client.
• Connect OpenRouter telemetry; send Broadcast traces to a public txtop receiver.
• Track Ollama usage; run the local proxy and add cost mappings.
• Inspect live agents; use and customize the Linux process view.
• Troubleshoot txtop; inspect collector, database, and source health.

Reference

• Configuration; fields, environment variables, defaults, paths, and precedence.
• Commands; TUI, collector, Ollama, and database command families.
• Keyboard; global navigation and dashboard bindings.
• Compatibility and support; the stable 1.x contract, experimental surfaces, platforms, and Rust version.
• Changelog; released behavior by version.
• Security policy; supported versions and private reporting.

Explanation

• Where the numbers come from; ingest paths, billing lanes, deduplication, and differences from provider totals.
• Unofficial integrations; the support and safety boundary for local parsers and live-limit requests.

To change txtop itself, use the contributor guide on GitHub. It covers architecture, extensions, testing, performance, and releases without mixing repository work into the product documentation.