Overview

DevoxxGenieIDEAPlugin adds an AI‑driven assistant directly into JetBrains IDEs such as IntelliJ IDEA, PyCharm, and WebStorm. The plugin is free, open‑source, and works with both local LLMs (Ollama, LMStudio, GPT‑4All, Jan, Llama.cpp) and cloud providers (OpenAI, Anthropic, Mistral, Groq, Kimi, GLM, Gemini, etc.). You install it from the JetBrains Marketplace and configure your own API keys, keeping the model selection under your control.

The core of DevoxxGenie is a Java‑based code assistant that can answer questions, suggest completions, and run tasks inside the IDE. Inline code completion uses Fill‑in‑the‑Middle models to generate context‑aware suggestions based on the code before and after the cursor. For larger tasks you can type “Create a task for …” and the plugin builds a structured spec with acceptance criteria, then hands it to the LLM agent for autonomous implementation. The spec‑driven workflow is backed by a built‑in Backlog.md tool that lets you browse, edit, and mark tasks as done without leaving the editor.

Agent Mode extends the assistant with read‑only tools that let the LLM explore the codebase, list directories, search patterns, and propose edits. You can run multiple sub‑agents in parallel, each configured with a different model, and you may keep the entire process local by using a model such as GLM‑4.7‑flash via Ollama. Advanced Model Context Protocol (MCP) support lets the assistant call external services, filesystem, databases, web browsing, through an MCP marketplace that can be browsed and installed from within the IDE. A debugging panel shows MCP requests and responses in real time, helping you understand how the model interacts with those tools.

Security scanning is built in: the assistant can run Gitleaks to find hard‑coded secrets, OpenGrep for static analysis, and Trivy for vulnerable dependencies. Findings are turned into prioritized backlog tasks automatically. The plugin also supports CLI runners, allowing you to execute prompts with external tools like Claude Code, GitHub Copilot, Codex, Gemini CLI, or Kimi directly from the chat window. If you work with multimodal models, you can drag and drop images into the chat for visual context.

Key features

Supports local LLMs (Ollama, LMStudio, GPT‑4All, Jan, Llama.cpp) and major cloud providers
Inline Fill‑in‑the‑Middle code completion with context from surrounding code
Spec‑driven development: create structured tasks with acceptance criteria and let the agent implement them

Details

Pricing: Free
Category: Code

Suggest an edit

Sign in to suggest an edit.

Pros and cons

Pros

Free and fully open source, no licensing fees
Works with a wide range of LLM providers, giving flexibility in model choice
Integrated task management and spec creation simplify AI‑assisted development
Local‑only operation possible, keeping data on your machine
Security scanning tools are bundled and generate actionable tasks automatically

Cons

Requires you to supply your own API keys for cloud models; no hosted key management is provided
Advanced features like MCP and Agent Mode may need additional configuration and understanding of the underlying protocols
Pricing information is not listed because the plugin is free, but costs for external LLM usage are not covered in the documentation.