Jinn Code

Multi-agent AI coding assistant for VS Code — powered by AWS Bedrock and local models.

Built by Jinn Systems · Sharjah, UAE

What is Jinn Code?

Jinn Code is a VS Code extension that brings a four-agent AI pipeline directly into your sidebar. Unlike single-model copilots, Jinn Code routes each part of a coding task to the model best suited for it — a Planner to break down the problem, a Thinker to reason through complexity, a Coder to write the implementation, and a Reviewer to catch issues before they hit your files.

It works with AWS Bedrock (Claude, Llama, Nova, DeepSeek) and any local model server (Ollama, LM Studio, or any OpenAI-compatible endpoint) — and you can mix providers per agent.

Features

• Three modes — Ask (Q&A), Plan (step-by-step breakdown), Agent (full pipeline with file writes)
• Four-agent pipeline — Thinker · Planner · Coder · Reviewer, each independently configurable
• Inline diff — AI-generated code appears as Keep/Discard hunks in your editor, Cursor-style
• Kernel Scheduler — route tasks to models by preset (Performance / Balanced / Custom)
• Live model fetch — pull your available Bedrock models or local server models directly into Settings
• Codebase indexing — keyword + semantic search over your workspace for relevant context injection
• MCP support — connect any Model Context Protocol server for extended tool use
• Export — save chat sessions as .txt or .json; copy debug logs for sharing
• Cross-platform — Windows, macOS, Linux

Quick Start

Requirements

• VS Code 1.85 or later
• Node.js 18 or later
• An AWS account with Bedrock access and/or a local model server (Ollama, LM Studio)

Install

Search "Jinn Code" in the VS Code Extensions marketplace and click Install.

Or install from the command line:

code --install-extension jinn-systems.jinn-code

Open the panel

Press Ctrl+Shift+J (Windows/Linux) or Cmd+Shift+J (macOS), or click the ✦ icon in the Activity Bar.

Setup

AWS Bedrock

1. Open the panel → click the 🔑 key icon in the header
2. Enter your AWS Access Key ID and Secret Access Key
3. Open Settings (⚙ gear icon) → AWS Bedrock tab
4. Set your region (e.g. us-east-1) and Auth Mode
5. Click Fetch Models to pull your available inference profiles
6. Go to General tab → assign models to each agent

Auth modes: | Mode | When to use | |------|-------------| | IAM / AWS CLI | Local dev with ~/.aws/credentials or instance role | | Access Key + Secret | Explicit credentials entered via the UI | | Endpoint URL | Bedrock Gateway or custom proxy |

Local Models (Ollama / LM Studio)

# Ollama
ollama pull llama3.1:8b
ollama serve

# LM Studio
# Start the local server from the LM Studio UI (default port 1234)

Then in Settings → Local Models tab:

1. Enter your endpoint URL (e.g. http://localhost:11434)
2. Click Fetch Models — your running models appear in a dropdown
3. Select a model and click Save Local Config

Modes

Ask

Single-agent Q&A. The Thinker reasons through your question with full workspace context. No file writes. Best for understanding code, debugging, and explaining concepts.

Plan

The Planner breaks your task into numbered steps. You review the plan before anything is written. Use Execute Plan to hand it to the Agent pipeline, Edit to refine in the prompt box, or Discard to start over.

Agent

Full pipeline. The Planner outlines the task, the Thinker reasons through complexity (triggered automatically for refactors, architecture changes, and migrations), the Coder writes the implementation, and the Reviewer checks it. Code appears as inline diffs in your editor — accept or discard each hunk individually or use Keep All / Discard All.

Agent Roles

Each agent can be independently assigned any provider and model via Settings → General.

Kernel Scheduler

The Kernel Scheduler decides which model handles each task type. Switch presets in Settings → Kernel:

Changes take effect on the next message — no reload required.

Inline Diff (Keep / Discard)

When the Coder writes to an existing file, changes appear as highlighted hunks in your editor:

• Green lines — AI's new content
• Red ghost text — what was there before
• ✔ Keep / ✖ Discard CodeLens buttons above each hunk
• Keep All / Discard All bar above the chat input

Clicking Keep or Discard automatically saves the file — no manual Ctrl+S needed.

Keyboard Shortcuts

MCP Servers

Jinn Code supports the Model Context Protocol. Add servers via settings.json:

"jinnCode.mcp": [
  {
    "name": "filesystem",
    "command": "npx",
    "args": ["-y", "@modelcontextprotocol/server-filesystem", "/path/to/project"]
  },
  {
    "name": "git",
    "command": "npx",
    "args": ["-y", "@modelcontextprotocol/server-git"]
  }
]

Servers start automatically when the extension activates.

Export & Logs

Chat export — click 📤 in the header:

• Copy to Clipboard — paste full chat as formatted text
• Save as Text (.txt) — timestamped text file
• Save as JSON (.json) — structured export with agent names, timing, and timestamps

Log panel — click 📄 in the header:

• Filter by level (Error / Warning / Info / Debug) or search by keyword
• 📋 Copy — copies a debug summary (errors first, then last 30 entries) for sharing
• Export logs as .txt or .json
• Red dot badge appears on the log icon when an error occurs

Requirements

• VS Code ^1.85.0
• Node.js 18+ (for fetch and AbortSignal builtins)
• AWS account with Bedrock model access enabled, or a running local model server

Privacy

Credentials are stored in VS Code's SecretStorage and never written to settings.json or disk. No telemetry is collected by this extension.

License

MIT © 2026 Jinn Systems