Continue Beans ☕

Enterprise-grade AI assistant for Apache NetBeans.

Continue Beans brings AI-assisted development directly into NetBeans — local and cloud LLMs, workspace context, tool calling, streaming chat, MCP integration, and native NetBeans language services.

---

Features

• AI chat integrated into NetBeans (dark, developer-focused UI)
• Multiple AI providers: LM Studio, Ollama, and any OpenAI-compatible API
• Streaming responses with live token updates
• Workspace & project context (files, directories, @file / codebase awareness)
• Tool / function calling with NetBeans APIs (read/write files, editor, projects)
• MCP (Model Context Protocol) tool integration
• Native NetBeans Java language services
• Conversation history and multi-turn context management
• Task orchestration and agent-style workflows
• Model discovery and configuration panel
• JUnit 5 + Cucumber test suite with JaCoCo coverage

---

Requirements

Requirement	Version / notes
Java	11+ (17/21 also supported)
Apache NetBeans	RELEASE200 (compatible with recent 20.x builds)
Maven	3.9+
AI backend	LM Studio, Ollama, or any OpenAI-compatible API

---

Installation

Option A — Install from release (recommended)

1. Download the latest .nbm from the Releases page.
2. In NetBeans: Tools → Plugins → Downloaded → Add Plugins… and select the .nbm file.
3. Install and restart NetBeans when prompted.
4. Open the assistant via Window → Continue Beans (or the equivalent menu entry).

Option B — Build from source

git clone https://github.com/offline0x33/continue-netbeans.git
cd continue-netbeans
mvn clean install

The generated NetBeans module is under target/*.nbm.

Run the full verification suite:

mvn clean verify

---

Quick start

1. Start an AI backend
   • LM Studio: load a model and enable the local server (usually http://localhost:1234).
   • Ollama: see Docker setup or run ollama serve locally.
2. Open NetBeans and the Continue Beans window.
3. Configure the provider URL / model in the options or chat controls.
4. Chat with workspace context enabled — the assistant can inspect project files and use NetBeans tools.

---

AI providers

Continue Beans uses a provider abstraction so backends can evolve independently from the UI.

Provider	Notes
LM Studio	Primary local provider; OpenAI-compatible API
Ollama	Local models; Docker helpers under docker/
OpenAI-compatible	Any server that speaks the OpenAI chat completions API

Configuration is managed through the NetBeans options panel and runtime model discovery.

---

Docker / Ollama

A ready-to-use Ollama stack (with qwen2.5:7b as a suggested model) lives in docker/.

cd docker
docker-compose up -d ollama
docker exec -it continue-beans-ollama ollama pull qwen2.5:7b

Full instructions: docker/README.md.

---

Architecture

NetBeans
   │
   ▼
Continue Beans UI (ChatPanel / TopComponent)
   │
   ├─ Conversation & streaming
   ├─ Workspace tools & file system
   ├─ MCP tools
   ├─ Task orchestrator
   └─ Native language services
           │
           ▼
       LLM Provider layer
           │
           ├─ LM Studio
           ├─ Ollama
           └─ OpenAI-compatible API

UI, providers, workspace tooling, and IDE integrations are separated so each layer can evolve independently.

---

User interface

Continue Beans ships a dark, developer-oriented chat as the canonical NetBeans view.

The UI covers conversation streaming, task progress, tool/status feedback, workspace context, model controls, warnings, message actions, and workspace status.

Visual contract: docs/UI_SPECIFICATION.md.

---

Testing

Stack: JUnit 5, Mockito, Cucumber, JaCoCo.

mvn test
mvn verify

---

Project status

Continue Beans is under active development.

Implemented:

• Core chat experience with streaming
• Workspace tooling and large-context handling
• Native NetBeans Java language services
• Function / tool calling with NetBeans APIs
• MCP integration
• Canonical dark theme entry point (ChatPanel)

Visual-fidelity refinements and additional agent workflows continue on the main line.

---

Documentation

Doc	Description
docs/	Technical docs, guides, and specs
docs/CHANGELOG.md	Notable changes
docs/UI_SPECIFICATION.md	UI visual contract
docs/LM_STUDIO_REAL_GUIDE.md	LM Studio setup
docker/README.md	Ollama Docker setup

---

Contributing

Contributions are welcome.

Before opening a pull request:

1. Build the project (mvn clean install).
2. Run the test suite (mvn test).
3. Run mvn verify.
4. Keep changes focused.
5. Add or update tests for new behavior.

---

License

Apache License 2.0.