SAGE Studio Agent — VS Code Extension

VS Code extension for SAGE Studio, providing a persistent AI agent workspace directly in your editor.

Features

• Chat Panel — multi-turn chat with SAGE Studio's LLM backend, streaming AgentStep responses with live thinking/tool-call indicators
• Agent Task Panel — submit natural-language tasks (research, coding, analysis) and watch the multi-agent swarm (ResearcherAgent, CodingAgent) work step by step
• Studio Backend Control — start and stop the sage-studio FastAPI backend from the Command Palette
• Status Bar — live connection indicator for sage-studio
• Context injection — highlight code, right-click → "SAGE Agent: Open Chat" to ask about the selection

Requirements

1. Python environment with SAGE installed:

   pip install isage isage-studio

2. Start the sage-studio backend (port 8765 by default):

   sage studio start

3. Optional — a running LLM endpoint (sageLLM gateway or OpenAI-compatible):

   sage-llm serve --model Qwen/Qwen2.5-7B-Instruct

Usage

Command	Description
SAGE Agent: Open Chat	Open the chat panel (streaming)
SAGE Agent: Open Agent Task Panel	Submit a multi-step agent task
SAGE Agent: Start Studio Backend	Run sage studio start
SAGE Agent: Stop Studio Backend	Stop the studio backend
SAGE Agent: Check Connection	Probe /health
SAGE Agent: Installation Guide	Show setup instructions

Extension Settings

Setting	Default	Description
sageAgent.studio.host	localhost	sage-studio backend host
sageAgent.studio.port	8765	sage-studio backend port
sageAgent.studio.tls	false	Use TLS for the connection
sageAgent.chat.systemPrompt	(default)	System prompt for the chat panel
sageAgent.studio.startCommand	sage studio start	Shell command to start the backend
sageAgent.studio.stopCommand	sage studio stop	Shell command to stop the backend

Architecture

sage-agent-vscode
├── src/
│   ├── extension.ts       ← activation, commands, process management
│   ├── studioClient.ts    ← HTTP/SSE client for sage-studio API
│   ├── chatPanel.ts       ← Chat webview (SSE stream → AgentStep rendering)
│   ├── agentPanel.ts      ← Agent Task webview (task submit + step feed)
│   └── statusBar.ts       ← Status bar indicator
└── package.json

The extension communicates with sage-studio at the following endpoints:

Endpoint	Usage
GET /health	Health check
POST /api/chat/v1/runs	Chat with SSE streaming
POST /api/agent/v1/runs	Agent task with SSE streaming

Development

git clone https://github.com/intellistream/sage-agent-vscode
cd sage-agent-vscode
npm install
npm run compile      # one-shot build
npm run watch        # watch mode

Press F5 in VS Code to launch the Extension Development Host.

Build & Package

npm run package             # production bundle → dist/extension.js
npx vsce package            # create .vsix

Related Projects

• sage-studio — the backend this extension connects to
• sagellm-vscode — SageLLM inference engine VS Code extension
• SAGE — SAGE meta-package

License

MIT © IntelliStream Team