EchoPilot is a revolutionary VS Code extension that transforms your development environment into a first-class AI workbench. Built on the Codessa AI platform, it provides intelligent code assistance, automated policy enforcement, and sophisticated agent-driven workflows through an intuitive chat interface, real-time diagnostics, and executable playbooks.
🚀 Vision: Bend VS Code into an AI-powered development workbench without forking the universe - leveraging existing Extension APIs to create a seamless AI development experience across all VS Code deployment scenarios.
- Smart Conversations: Chat with AI about your code using
@codessain VS Code's chat panel - Contextual Awareness: Automatically includes workspace context, open files, and selected text
- Streaming Responses: Real-time AI responses with markdown formatting
- Direct Actions: AI can create, edit, and delete files directly in your workspace
- Command Execution: Run terminal commands suggested by the AI
- Real-time Policy Checking: Automatic policy validation on file save
- Inline Diagnostics: Policy violations shown as VS Code diagnostics with severity levels
- Quick Fixes: One-click fixes for common policy violations
- Code Actions: "Fix all" and "Explain policy" actions in the editor
- Custom Rules: Support for organization-specific GitGuard policies
- Notebook Interface: Define multi-step AI workflows as executable notebooks
- Step Types: Plan, Search, Edit, Test, and Analyze steps
- Visual Execution: See each step's progress and output in real-time
- Reusable Workflows: Save and share common automation patterns
- JSON/YAML Configuration: Flexible step definitions
- vscode.dev Compatible: Works in browser-based VS Code environments
- Codespaces Ready: Full functionality in GitHub Codespaces
- Remote Development: Supports SSH, containers, and remote workspaces
- Fallback Modes: Graceful degradation for limited web environments
- Lightweight Satellite: Embed Monaco editor in web applications
- AI-Powered: Full Codessa AI assistance in embedded editors
- Policy Checking: Real-time policy validation in Monaco
- Customizable: Configurable themes, languages, and features
- VS Code: Version 1.85.0 or higher
- Node.js: Version 20.x or higher
- Git: For version control and collaboration
- Codessa API Key: Sign up at codessa.dev for API access
# Install via VS Code Extensions view
1. Open VS Code
2. Press Ctrl+Shift+X (Cmd+Shift+X on Mac)
3. Search for "EchoPilot"
4. Click "Install"# Clone and setup for development
git clone https://github.com/username/EchoPilot.git
cd EchoPilot
npm install
npm run compile
# Launch Extension Development Host
code . && press F5# Download and install .vsix package
npm install -g vsce
vsce package
code --install-extension echopilot-*.vsix-
Get API Credentials
- Visit codessa.dev and create an account
- Generate an API key from your dashboard
- Copy your API endpoint URL
-
Configure VS Code Settings
{ "codessa.apiEndpoint": "https://api.codessa.dev", "codessa.apiKey": "your-api-key-here" } -
Optional Advanced Settings
{ "codessa.enablePolicyChecks": true, "codessa.streamResponses": true, "codessa.webApiEndpoint": "https://web-api.codessa.dev", "codessa.maxTokens": 4096, "codessa.temperature": 0.7 }
{
"codessa.webApiEndpoint": "https://web-api.codessa.dev",
"codessa.corsEnabled": true
}{
"codessa.apiEndpoint": "https://your-domain.com/api",
"codessa.customHeaders": {
"X-Organization": "your-org-id"
}
}- Open VS Code's chat panel (Ctrl+Alt+I)
- Type
@codessafollowed by your question - The AI will respond with contextual help and can perform actions
Example conversations:
@codessa explain this function(with code selected)@codessa refactor this component to use hooks@codessa create a test file for this module@codessa fix the TypeScript errors in this file
- Automatic: Policies are checked when you save files
- Manual: Use Command Palette → "Codessa: Check GitGuard Policies"
- Quick Fixes: Click the lightbulb icon on policy violations
- Use Command Palette → "Codessa: Run Agent Playbook"
- A new notebook opens with example steps
- Modify the JSON configuration for each step
- Execute cells to run the workflow
Example Playbook Steps:
{
"type": "plan",
"description": "Analyze codebase for refactoring opportunities",
"input": {
"task": "Identify components that could be converted to hooks",
"scope": "src/components"
}
}- Use Command Palette → "Codessa: Open Monaco Editor"
- Or "Codessa: Open Current File in Monaco Editor"
- Embedded editor with full AI assistance
The extension works seamlessly in web environments:
- Go to vscode.dev
- Install the Codessa AI Workbench extension
- Configure your API settings
- Start using AI assistance in the browser
- Create or open a Codespace
- The extension will be available if pre-installed
- All features work with some limitations for file system operations
src/
├── extension.ts # Main extension entry point
├── api/
│ └── apiClient.ts # Codessa API communication
├── chat/
│ └── chatParticipant.ts # @codessa chat participant
├── policy/
│ └── diagnosticsProvider.ts # GitGuard policy checking
├── notebook/
│ └── playbookController.ts # Agent playbook execution
├── monaco/
│ └── monacoIntegration.ts # Monaco editor integration
└── web/
└── webExtension.ts # Web compatibility layer
- Chat Participant: Handles
@codessamentions and routes to AI backend - Policy Provider: Integrates GitGuard with VS Code diagnostics system
- Notebook Controller: Executes agent playbooks as interactive notebooks
- API Client: Manages communication with Codessa platform
- Web Compatibility: Ensures functionality across all VS Code environments
- Node.js ≥ 20
- VS Code ≥ 1.85.0
- TypeScript
git clone https://github.com/codessa/vscode-extension
cd vscode-extension
npm install
npm run watchnpm run testnpm run compilenpm install -g vsce
vsce packageThe extension communicates with the Codessa platform through REST APIs:
POST /chat
{
"message": "user prompt",
"context": {
"workspaceRoot": "/path/to/workspace",
"activeFile": "src/component.tsx",
"selectedText": "selected code"
},
"stream": true
}POST /policy/check
{
"filePath": "src/component.tsx",
"content": "file contents",
"language": "typescript"
}POST /playbook/execute
{
"id": "step-123",
"type": "edit",
"description": "Refactor component",
"input": { "files": ["src/component.tsx"] }
}- Fork the repository
- Create a feature branch
- Make your changes
- Add tests
- Submit a pull request
MIT License - see LICENSE file for details.
- Documentation: docs.codessa.dev
- Issues: GitHub Issues
- Discord: Codessa Community
- Email: support@codessa.dev
Transform your VS Code into an AI-powered development workbench with Codessa! 🚀