See how many messages and sessions you have used across every AI coding tool.
Track estimated spend across different AI editors and models.
Compare two AI coding tools side by side to see usage differences.
Share coding session context with a team using the Relay feature.
| f/agentlytics | sveltejs/rollup-plugin-svelte | typicode/fetchival | |
|---|---|---|---|
| Stars | 518 | 516 | 520 |
| Language | JavaScript | JavaScript | JavaScript |
| Last pushed | — | 2026-05-20 | 2018-12-06 |
| Maintenance | — | Maintained | Dormant |
| Setup difficulty | easy | easy | easy |
| Complexity | 2/5 | 2/5 | 2/5 |
| Audience | developer | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
macOS only, Linux and Windows are not yet supported.
Agentlytics is a local analytics dashboard for developers who use AI coding assistants. If you switch between tools like Cursor, Windsurf, Claude Code, VS Code Copilot, and others, each of those tools stores your conversation history separately with no shared view. Agentlytics reads all of those local files and brings the data together into one place. You run it with a single command, and it opens a web dashboard in your browser on your own computer. Nothing is uploaded anywhere. The dashboard shows you how many sessions you have had, how many messages, which AI models you used most, what tools were called during those sessions, estimated costs based on token counts, peak usage hours, and coding streaks. It covers 17 editors at the time of the README. The cost tracking feature is useful for understanding how much you are spending across different editors and models. The compare view lets you put two editors side by side to see which one you use differently. There is also a subscriptions view that shows your current plan limits and remaining credits for the editors that expose that information. A feature called Relay allows a team to share session context with each other. One person starts a relay server on their machine, others join it, and an AI client can then query across the whole team's coding history. This uses a protocol called MCP for the AI client connection. The project currently supports macOS only, Linux and Windows support is listed as a future goal.
A local dashboard that combines your usage history from 17+ AI coding tools like Cursor and Claude Code into one view of sessions, costs, and models used.
Mainly JavaScript. The stack also includes JavaScript, Node.js, MCP.
No license information was found in the material provided.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.