MCP server + Chrome extension. Your agent controls your REAL browser (same tabs, cookies, logins). No headless browser, no replaying auth. One-click install in Cursor, extension on Chrome Web Store.Repo: https://github.com/ofershap/real-browser-mcp Chrome extension: https://chromewebstore.google.com/detail/real-browser-mcp/fk...
I built this because I kept wasting time on README formatting -- badge URLs, table alignment, pushing just to check rendering.It's a client-side React app. You drag blocks (headings, badges, code, tables, images, alerts) into a canvas and see exactly how GitHub will render them. When it looks right, copy or export.No backend, no login. Source: https://github.com/ofershap/readme-builder
Built a retro TUI that connects to the Cursor Agents API - shows all agents on one screen, color-coded by status. You can launch new agents, send follow-up prompts, and stop runaway ones without leaving the terminal.
npx agents-control-tower to try it. MIT, TypeScript, zero config.
Turns out half my .cursor/rules were vague stuff like "follow best practices" that just wasted tokens. And my CLAUDE.md overlapped with AGENTS.md on half the instructions.
So I built a small tool that measures the token cost per file and lints for conflicts/dead weight: https://github.com/ofershap/ai-context-kit
Run `npx ai-context-kit measure` on any project and you get a per-file token breakdown. Helped me cut my context budget by ~40% without losing anything useful.