Pythinker Code

AI coding assistant for VS Code, built for long-context workflows and complex coding tasks.

Features

• Works alongside you: Pythinker autonomously explores your codebase, reads and writes code, and runs terminal commands with your permission
• Thinking controls: Toggle reasoning or choose a model-supported thinking effort
• Provider-aware models: Distinguish and select same-named models across configured providers
• Native editor integration: Review AI-proposed changes directly in VS Code's diff viewer
• MCP support: Extend capabilities with Model Context Protocol servers
• Slash commands: Quick actions like /init to analyze your project and /compact to manage context

Install

Pythinker Code requires VS Code 1.100.0 or later.

1. Install from the VS Code Marketplace
2. Open a folder in VS Code
3. Click the Pythinker icon in the Activity Bar
4. Sign in, or use a provider already configured in the shared config.toml

The extension runs the Pythinker Code Node SDK in the VS Code Extension Host. When the extension and the Pythinker Code terminal app resolve to the same PYTHINKER_CODE_HOME, they share config.toml, MCP configuration, login state, and sessions. The system-level PYTHINKER_CODE_HOME environment variable is supported; there is no separate VS Code setting for it. Do not run the same session from both applications at the same time, because cross-process session locking is not guaranteed.

Extension development

The F5 launch profile ("Extension Development Host (isolated)") runs against a disposable PYTHINKER_CODE_HOME at .tmp/vscode-extension-dev/pythinker-home, recreated empty on every launch by scripts/prepare-dev.mjs. It does not read your real ~/.pythinker-code/config.toml.

To start the dev extension with your real providers and models, use the "(seeded config)" launch profile or pnpm run dev:prepare:seeded. It copies your real config.toml verbatim into a disposable dev home, which duplicates every secret it contains — API keys and any embedded provider auth metadata — into a file under .tmp/ (mode 0600 where supported). Credential stores, session data, and mcp.json stay in the real home.

The seeded profile lives in its own tree (.tmp/seeded/vscode-extension-dev) with its own user-data and extensions directories, so the isolated profile never inherits the seeded run's configuration or stored credentials.

Note: the engine watches config.toml and reloads itself on change, but the extension has no config-change subscription. The model picker and Models section render Zustand state — no RPC — so they update only on webview reinitialization or extension-owned provider changes that emit ProvidersChanged. The raw Config File tab rereads the TOML display when opened.

Docs

Full documentation is available at pymodel.github.io/pythinker-code.

License

Apache-2.0