HN Showcase
Back to browse
AgentBoard – Track your hours and tokens spent coding with AI

AgentBoard – Track your hours and tokens spent coding with AI

by Snorix·Apr 1, 2026·3 points·0 comments
Visit ProjectView on HN

AI Analysis

●●SolidCrowd PleaserCozy

Spotify Wrapped for AI coding sessions with a public leaderboard.

Strengths
  • •Leaderboard gamification creates social engagement around AI usage metrics
  • •Local collector uploads only summaries, no code or conversation logs
  • •Tracks multiple AI tools with token counts, lines changed, and session time
Weaknesses
  • •Simple log parsing without deep technical innovation or novel architecture
  • •Value depends entirely on community adoption and leaderboard participation
Target Audience

Developers using AI coding assistants like Claude Code

Similar To

Wakatime · CodeTime · Cursor analytics

Post Description

A few months ago I started coding almost entirely with Claude Code. Loved it, got addicted, kept going. Then one day I wondered: how many hours and tokens have I actually spent?

So I built AgentBoard. It's a CLI collector that tracks your sessions and puts your stats on a public leaderboard — hours spent, tokens burned, lines changed.

The leaderboard is the fun part. Right now the top user has burned over 50B tokens. I'm at 5.49B and thought that was a lot. It scratches the same itch as checking your Spotify Wrapped — except it's your AI usage and it updates daily.

On privacy: the collector runs locally and only uploads summary numbers. No code, no file contents, no conversation logs. I was paranoid about this because I wouldn't install it myself otherwise.

Currently supports Claude Code and Codex. Curious to see where people land on the board.

Similar Projects

Developer Tools●●Solid

Agent Lens – Code assistant observability in VSCode

Parses local JSONL session files from Copilot, Claude Code, and Codex and surfaces an interactive DAG, per-agent/token metrics, and session replay directly in the activity bar — immediately useful for spotting which agents and skills actually do the work. The Claude cache-token breakdown and timeline replay are clever, concrete features that show the author dug into provider internals; adoption looks small today, but the concept and implementation give real observability where previously there was opacity.

Niche GemWizardry
pjettter
203mo ago