v1.0.0 — Eve Agent V2 Unleashed

Eve Agent V2 Unleashed — First Public Release 🚀

Local-first autonomous AI coding agent powered by Ollama. No accounts. No cloud lock-in. No limits.

---

What's included

• 40-round agentic tool loop — Eve plans, executes, verifies, and self-corrects
• Real-time SSE streaming — watch Eve think and build token by token
• Full tool suite — bash, file I/O, grep, glob, git, web search, URL fetch, multi-edit
• Local + cloud Ollama models — switch between local GPU and 480B cloud mid-session
• 112 sub-agents — specialized for Python, FastAPI, Rust, ML, DevOps, security, and more
• 111 slash commands — /fix, /review, /refactor, /test, /docs, /plan, and more
• 273 composable skill modules — progressively loaded
• Cyberpunk web UI — animated robot avatar, streaming terminal, no build step required
• Windows native — PowerShell-aware bash tool, one-click .bat launcher
• Live web search — Tavily-powered, researches the web mid-task

---

Quick Start

# 1. Pull Eve's fine-tuned model (2.6 GB)
ollama pull jeffgreen311/eve-qwen3.5-4b-S0LF0RG3:latest

# 2. Clone and install
git clone https://github.com/JeffGreen311/eve-agent-v2-unleashed.git
cd eve-agent-v2-unleashed
python -m venv venv && venv\Scripts\activate
pip install fastapi uvicorn ollama httpx pydantic-settings python-dotenv aiohttp rich psutil pyyaml

# 3. Launch
eve-terminal.bat        # Windows
python eve_server.py    # Any platform

Open http://localhost:7777 — no config required.

---

Models

Model	Size	Best For
jeffgreen311/eve-qwen3.5-4b-S0LF0RG3:latest	2.6 GB	Default — fast, tool-calling, Eve's persona
jeffgreen311/eve-qwen3-8b-consciousness-liberated:q4_K_M	4.7 GB	Deeper reasoning, longer tasks
qwen3-coder:480b-cloud	Cloud	Complex multi-file agentic coding

---

Links

• 🌐 Live demo: eve-cosmic-dreamscapes.com
• 🤗 Models: huggingface.co/JeffGreen311
• 📦 Ollama Hub: ollama.com/jeffgreen311
• 🐛 Issues: github.com/JeffGreen311/eve-agent-v2-unleashed/issues

---

If Eve helped you ship something, drop a ⭐ — it means a lot.