ccusage and OpenLimiter
The broadest local log reader, and the only genuinely cross platform tool here.
- Platform
- Command line, Node based, run with npx. Windows, macOS and Linux.
- Licence
- MIT
- Coverage
- Claude Code, Codex, OpenCode, Amp, Droid, Codebuff, Hermes Agent, pi-agent, Goose, OpenClaw, Kilo, Kimi, Qwen, GitHub Copilot CLI and Gemini CLI, as its site lists them.
- Source
- Local files only. It uploads nothing and calls no provider API.
What it does
It reads the JSONL session logs that coding agent command line tools already write to your disk and turns them into daily, weekly, monthly and per session token and cost reports. Its own description is analysing coding agent token usage and costs from local data.
What it does well
The breadth of its log parsers is unmatched: more than fifteen agent tools from one binary, with no credentials and no setup at all. It also emits structured JSON and ships a Claude Code statusline hook.
Where OpenLimiter differs
It reports what you already spent, not what you have left. Its only quota notion is a token limit you supply yourself on the blocks report, so it never reads a real remaining figure from a provider. OpenLimiter starts from the provider's own remaining percentage and hands that to an agent as a bounded block.
How these were checked
These notes were read off each project's own repository, README or store listing in August 2026. Software moves, so check the source before you rely on any of it, and open an issue if something here has gone stale or is simply wrong.