I built a local-first autonomous coding agent with a cyberpunk soul — Eve Agent V2 Unleashed (open source)

2026-05-29 · Source: Artificial Intelligence · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Robotics & Autonomous Systems · Depth: Intermediate, quick

Summary

Eve Agent V2 Unleashed is an open-source, local-first autonomous coding agent designed to run on a user's local GPU via Ollama, offering optional cloud escalation for complex tasks. This agent can autonomously plan, write, test, debug, and verify code for tasks like building a FastAPI server with JWT auth and a PostgreSQL backend. It features a 40-round agentic loop utilizing 112 specialized sub-agents and 273 composable skill modules, with tools including bash, file I/O, grep, git, web search, multi-edit, and computer vision via OpenClaw. The agent demonstrated 9/9 tests passing on its first attempt with zero intervention using "qwen3-coder:480b-cloud" and a local 8B Q4_K_M model. It includes 111 slash commands, an RPG progression system, a Telegram bridge, and Windows native support, emphasizing no accounts or telemetry. Fine-tuned models like "jeffgreen311/Eve-V2-Unleashed-Qwen3.5-8B-Liberated-4K-4B-Merged" (3.4 GB) are available on Ollama Hub for quick setup.

Key takeaway

For AI Engineers or Software Developers seeking to automate coding tasks, Eve Agent V2 Unleashed offers a robust, local-first solution. You can deploy this open-source agent on your local GPU using Ollama to handle complex development cycles from planning to debugging. This approach minimizes reliance on cloud services and ensures data privacy. Consider integrating Eve Agent V2 into your workflow to offload repetitive coding, leveraging its extensive toolset and autonomous capabilities.

Key insights

Eve Agent V2 is an open-source, local-first autonomous coding agent that uses a multi-round agentic loop and diverse tools for complex software development tasks.

Principles

• Autonomous agents can perform complex coding tasks end-to-end.
• Local-first LLM execution is viable for sophisticated development.
• Multi-tool integration enhances agent capabilities significantly.

Method

Eve Agent V2 employs a 40-round agentic loop, planning, writing, testing, and debugging code using 112 sub-agents and 273 skill modules, integrating tools like bash, Git, web search, and computer vision.

In practice

• Deploy Eve Agent V2 locally via Ollama for autonomous coding.
• Utilize its 111 slash commands for specific development actions.
• Integrate the Telegram bridge for mobile task notifications.

Topics

• Autonomous Agents
• Local LLMs
• Code Generation
• Ollama
• Open-Source Software
• Computer Vision
• Software Development Automation

Code references

• JeffGreen311/eve-agent-v2-unleashed
• JeffGreen311/eve-agent-v2-unleashed

Best for: Computer Vision Engineer, AI Engineer, Machine Learning Engineer, Software Engineer

Related on AIssential

• Open Source Coding Agents: The Future of AI-Driven Software Development — Open source coding agents enable autonomous, scalable software development by orchestrating multi-agent workflows and offering model flexibility.
• Codex + Ollama = Free Unlimited Coding AI — Ollama's integration with Codex enables free, local execution of open-source LLMs for AI coding tasks.
• You Asked About AI: Agents, Hacking & LLMs — AI agents are democratizing complex tasks like hacking and enabling new forms of inter-agent communication.
• Run Qwen3.5 on an Old Laptop: A Lightweight Local Agentic AI Setup Guide — Local AI setups are accessible on older hardware using lightweight models and tools like Ollama and OpenCode.
• OpenCode Supercharged: BEST AI Coding Agent Ever + FULLY FREE! BYE Gemini CLI & ClaudeCode! — Oh My Open Code transforms basic AI coding agents into robust, multi-agent development teams with advanced orchestration.
• Y Combinator-backed Random Labs launches Slate V1, claiming the first 'swarm-native' coding agent — Slate V1 introduces "swarm-native" agentic coding via "Thread Weaving" for scalable, multi-model software development.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.