· Open Source · Free Forever

You hit your AI limit.
You lose context. Again.

One command. Context copied. Paste into Claude.ai. Resume in seconds.

Which fits you?

Just need context recovery?

Cursor, Claude Code, Copilot — no setup. Run, paste, resume.

npx github:abhishek5878/devflow

Need proxy routing?

Continue.dev, Roo Code, Cline — VS Code extension, one-click connect, auto-failover.

Get Extension

100% local Zero config No data sent GitHub

Three steps. Ten seconds.

No install. No sign-up. Your data never leaves your machine.

In your project directory. That's it.

Copies git history, stack, rules, errors to your clipboard.

Resume exactly where you left off. No re-explaining.

Cursor, Claude Code, Copilot users — Hit a limit? Run the command, paste, resume. No proxy needed.

Continue.dev, Roo Code, Cline users — VS Code extension: one-click connect, auto-failover, token dashboard.

Proxy adds Claude → GPT-4 → Gemini → Ollama. Run ollama run llama3.2 for free local fallback when paid limits hit.

Session context — Changed files, recent commits

Stack + rules — Next.js, TypeScript, .cursor/rules

Known errors — TypeScript, terminal—next AI hits ground running

Open a terminal in your project. Paste the command. Done.

First run may take ~30s (npx installs deps).

npx github:abhishek5878/devflow