Hey PH! I'm Vick, a full-stack developer and solo maker. I built ai-spend-tracker because I was juggling Claude Code, Hermes, Codex, OpenCode, Copilot, OpenClaw, and Kimi every day — and had NO idea how much I was spending across all of them. Each tool has its own billing, none of them talk to each other, and the costs add up silently. So I built a CLI that reads the local logs from 7 different AI coding agents, normalizes their wildly different data formats into one standard model, and gives you a clean report: tokens used, estimated cost, by agent and by model. It's open source (MIT), cross-platform, and built with a plugin architecture — adding a new agent is just writing one collector. I'm a frontend developer who learned Python backend while building this (with heavy AI assistance). This project went from idea to PyPI release in 3 months, solo. Would love feedback from other devs using multiple AI coding tools: - What other agents should I support? - How do YOU track your AI tool costs? - What would make this a daily-use tool for you? Thanks for checking it out! 🚀

AI Spend Tracker

oken-usage , cost-tracking , ai-coding-agent

oken-usage , cost-tracking , ai-coding-agent