A new open-source analytics platform called Agentlytics addresses a growing problem for developers: tracking AI coding assistant usage across multiple tools. Released on GitHub on March 7, 2026, the project has gained 306 stars and delivers a unified dashboard for 16 different AI coding editors and assistants, all while keeping data local on users' machines.
Single Dashboard Tracks Cursor, Windsurf, Claude Code, and 13 Others
Agentlytics supports a comprehensive range of AI coding tools including Cursor, Windsurf and Windsurf Next, Claude Code, VS Code with Copilot, Zed, OpenCode, Codex, Gemini CLI, Antigravity, Command Code, Goose, and Kiro. The platform's tagline captures its core value proposition: "One command to turn scattered AI conversations from 16 editors into a unified analytics dashboard."
The tool addresses fragmentation in the AI coding space, where developers often switch between multiple assistants depending on the task, project, or available features. Previously, tracking usage patterns and costs across these tools required checking each service individually.
Comprehensive Metrics Cover Costs, Sessions, and Productivity Patterns
Agentlytics provides detailed analytics across multiple dimensions:
- Session tracking across all editors in a unified view
- Cost estimation by model and editor
- Token usage breakdown
- Project-level analytics
- Tool usage frequency
- Daily activity heatmaps
- Coding streaks and peak productivity hours
- Subscription status and rate limits across services
These metrics give developers visibility into which tools they use most, which are most cost-effective, and how their AI-assisted coding patterns change over time.
Local-First Architecture Keeps Data Private
Unlike cloud-based analytics services, Agentlytics emphasizes a "local-first" architecture. All data stays on the user's machine rather than being uploaded to external servers. The implementation uses Node.js, React, and SQLite for data storage, making it possible to analyze sensitive coding sessions without privacy concerns.
This architectural choice aligns with growing developer concerns about sharing code and proprietary information with third-party services, especially as AI coding tools become more prevalent in enterprise environments.
Part of Growing Demand for AI Coding Analytics
Agentlytics emerged alongside other analytics tools like Rudel, which focuses specifically on Claude Code sessions. The simultaneous development of multiple analytics platforms suggests growing demand for visibility into AI coding agent usage patterns, costs, and effectiveness. With 167 commits already in the repository, the project shows active development and community engagement.
Key Takeaways
- Agentlytics provides unified analytics for 16 AI coding tools including Cursor, Windsurf, Claude Code, VS Code Copilot, and Zed
- The platform uses a local-first architecture, storing all data on users' machines rather than uploading to cloud services
- Analytics include cost estimation, token usage, session tracking, productivity patterns, and rate limit monitoring
- The project has gained 306 GitHub stars since launching on March 7, 2026, with 167 commits showing active development
- Growing number of AI coding analytics tools suggests increasing demand for visibility into AI assistant usage and costs