whoburnedmore

Product Introduction

1. Definition: whoburnedmore is a free, open-source command-line interface (CLI) tool and cross-platform dashboard that serves as a centralized analytics hub for tracking and comparing AI coding-agent usage metrics. It functions as a meta-aggregator, parsing local usage data from a wide array of popular AI development tools.
2. Core Value Proposition: The product provides developers with instant, comprehensive visibility into their AI-assisted coding consumption—tokens, estimated costs, tool breakdowns, and model usage—across 14+ platforms in a single unified interface. Its primary keywords are AI coding usage dashboard, token cost tracker, and cross-tool leaderboard.

Main Features

1. Universal Cross-Platform Dashboard: The tool runs a single command (npx whoburnedmore) to instantly generate a local dashboard. It automatically scans and aggregates historical and real-time usage data from 14 different AI coding assistants, including Claude Code, Codex, Cursor, Open Code, GitHub Copilot, Amp, Droid, Goose, Kimi, Qwen, and more. Technically, it reads local configuration and log files from each supported tool, normalizes disparate data formats, and calculates totals for tokens consumed, estimated financial cost, per-tool breakdown, and streak metrics. This eliminates the need to manually check each tool's native usage interface.
2. Public & Competitive Leaderboard: After generating your dashboard, the tool uploads only your aggregate, daily statistical totals (never your code or prompts) to a live, public leaderboard. Users can rank themselves against a global community of developers, comparing metrics like total tokens burned, estimated cost, daily burn rate, and top tools/models used. The ranking can be filtered by tokens, cost, all-time stats, or specific tools like claude, codex, cursor, and opencode.
3. Fully Local & Private Mode (--local): Addressing significant privacy concerns, the tool offers a --local flag (npx whoburnedmore --local). When used, it generates and serves the complete dashboard locally on the user's machine. In this mode, absolutely no data leaves the user's device, providing a completely private analytics experience for sensitive or corporate environments.
4. Custom Leaderboard Creation: The platform allows users to create and manage custom leaderboards, enabling teams, organizations, or communities to track and benchmark AI usage among a select group, separate from the global public board.

Problems Solved

1. Pain Point: The primary problem is the fragmentation and lack of visibility in AI coding costs and usage. Developers using multiple AI tools (e.g., Cursor for IDE integration, Claude Code for terminal tasks, Codex for specialized code) have no consolidated way to understand their total consumption, spending, or comparative efficiency, leading to unexpected expenses and opaque workflows.
2. Target Audience:
   • Individual AI Power Users: Developers heavily reliant on multiple AI coding assistants who need to track personal usage patterns and costs.
   • Engineering Managers & Team Leads: Professionals who need to benchmark team usage, justify AI tool budgets, and identify cost-saving opportunities across developers.
   • Open Source Contributors & the AI-Curious Community: Users interested in comparing their AI utilization against public trends and learning which tools and models are most popular.
   • Data Analysts & FinOps Roles: Those responsible for tracking and optimizing software development expenditures, including cloud and AI service costs.
3. Use Cases:
   • Monthly Cost Auditing: A developer runs the tool to see a consolidated monthly report of all AI tool costs to manage their budget.
   • Tool Adoption Analysis: A tech lead uses the dashboard to see which AI tools their team actually uses, informing future toolchain investments.
   • Community Benchmarking: A contributor checks their token usage against the public leaderboard to gauge their utilization against industry peers.
   • Corporate Privacy Compliance: A developer in a secure environment uses the --local flag to analyze usage without violating data governance policies.

Unique Advantages

1. Differentiation: Unlike native usage dashboards from individual providers (e.g., Anthropic's dashboard for Claude, OpenAI's for Codex), whoburnedmore is vendor-agnostic and aggregative. It breaks down silos by providing a single-pane-of-glass view. Compared to manual spreadsheet tracking, it offers automation, historical tracking, and community comparison as core features. Its open-source nature (built on ccusage MIT library) also allows for full inspection and community trust.
2. Key Innovation: The core innovation is its privacy-by-design architecture coupled with community gamification. The technical approach of only uploading daily aggregate metrics (token counts, cost estimates) while leaving all sensitive code, prompts, and file names strictly on the local machine is a critical differentiator. Furthermore, integrating this private analytics with a competitive, social leaderboard creates a novel ecosystem that encourages responsible usage tracking and community engagement.

Frequently Asked Questions (FAQ)

1. How does whoburnedmore keep my code and prompts private? The tool is designed with privacy as a core principle. It only reads local usage log files to calculate aggregate statistics. When you run the standard command, only those daily aggregate totals (e.g., "you used 10,000 tokens today") are uploaded to the public leaderboard. For absolute privacy, using the --local flag ensures no data whatsoever leaves your machine, and the dashboard runs entirely locally.
2. Which AI coding tools are supported by whoburnedmore? The tool currently supports over 14 platforms, including Claude Code, Codex, Cursor, GitHub Copilot, Open Code, Amp, Droid, Goose, Kimi, Qwen, and is actively expanding its integrations. It provides a unified dashboard for all these disparate tools.
3. Is there a cost to using whoburnedmore or its leaderboard? No, the tool and the public leaderboard are completely free to use. It is an open-source project. The "cost" data displayed is an estimate of API costs based on your token usage, not a charge from whoburnedmore.
4. Can I create a private leaderboard for just my team? Yes. After generating your dashboard, you have the option to "Make a custom leaderboard." This feature allows you to set up a separate, invite-only board for tracking and comparing usage within a specific group, team, or organization.
5. What specific usage metrics does the dashboard track? The dashboard tracks total tokens burned, estimated cost, daily and weekly usage, usage streaks, breakdown by tool (e.g., % from Claude vs. Cursor), top models used, and provides a detailed, sortable table for all these metrics across supported platforms.