Wayne

An autonomous coding agent in your IDE.

Wayne creates and edits files, runs commands, and works through multi-step engineering tasks — asking for your approval at every step.

What it does

Plans before it codes. Wayne breaks a request into a plan, asks clarifying questions, and shows you its approach before touching a file.

Understands your codebase. It reads the relevant files, follows imports and call sites, and makes changes that match your existing conventions.

Runs commands. Wayne executes terminal commands, reads the output, and reacts to failures — installing packages, running tests, and fixing what breaks.

Edits with diffs you approve. Every change is presented as a reviewable diff. Nothing is written without your say-so.

Extends through MCP. Connect Wayne to databases, internal APIs, and other services using the Model Context Protocol.

Getting started

1. Open the Wayne panel from the activity bar.
2. Open Settings and add an API key for your provider (Anthropic, OpenAI, Google, and others are supported).
3. Describe what you want built.

Controls

Wayne asks permission before writing files or running commands. You can approve each action individually, or enable auto-approve for the operations you trust.

Built by Samir Sengupta.