Tessa Agent is your intelligent AI coding assistant for Visual Studio Code, powered by OpenAI-compatible APIs.
- Tabbed Sidebar Interface:
- Chat Tab: Interactive chat with Tessa Agent.
- Context-aware responses using your active file and selection.
- Supports markdown for rich responses.
- Refresh Context: Button to manually update Tessa with the latest editor state.
- Insert Code: Easily insert AI-generated code snippets into your editor at the cursor.
- Edit Messages: Modify your previous prompts and Tessa will re-evaluate the conversation.
- Chat Commands:
/fim: Perform Fill-In-Middle at your current cursor position./explain: Ask Tessa to explain selected code or the current file./update_file <instruction>: Request Tessa to propose changes to the active file (review via diff).
- Settings Tab:
- Configure multiple AI models (name, API model ID, specific API URL/Key, type).
- Assign your configured models to different use cases (Chat, Inline Completion, FIM).
- Chat Tab: Interactive chat with Tessa Agent.
- Inline Code Completion: Get AI-powered code suggestions as you type.
- Configurable: Set your global OpenAI API Key and URL in VS Code settings. Override per model in the sidebar settings.
- Clone this repository.
- Run
npm installin the root directory. - Run
npm run compile(ornpm run watch). - Open the project in VS Code.
- Press
F5to launch the Extension Development Host.
Required:
- In the Extension Development Host window, open VS Code Settings (Ctrl/Cmd + ,).
- Search for "Tessa Agent".
- Set your global
Tessa Agent: Api Key.
Optional (via VS Code Settings):
Tessa Agent: Api Url: If using a proxy or custom endpoint.Tessa Agent: Model Assignment: Set default model IDs for Chat, Completion, and FIM. These IDs must correspond to models you configure in the sidebar.Tessa Agent: Model Configurations: Pre-populate some common models. You can manage these more easily via the "Settings" tab in the Tessa Agent sidebar.
Via Tessa Agent Sidebar "Settings" Tab:
- Add/Remove Models: Define custom names, API model IDs, specific API URLs/Keys (optional, overrides global), and model types (chat/completion).
- Assign Models: Choose which of your configured models should be used for Chat, Inline Completion, and Fill-In-Middle tasks.
- Open Tessa Agent: Click the Tessa Agent icon in the VS Code activity bar.
- Chat: Use the "Chat" tab. Type messages or commands.
- Configure: Use the "Settings" tab to manage your AI models.
- Completion: Start typing in your code files.
- FIM: In the chat, type
/fimand Tessa will use the code around your current cursor to fill in the middle.
