{
  "id": "local-private-coding",
  "title": "Local, private coding agent",
  "category": "coding-agents",
  "job": "Use an AI coding agent on code that must never leave your machine.",
  "difficulty": "Intermediate",
  "tools": [
    {
      "id": "ollama",
      "name": "Ollama",
      "url": "https://ollama.com",
      "kind": "local-runtime",
      "summary": "Run open models locally. Has one-command launchers for Claude Code and OpenCode.",
      "open_source": true,
      "pricing": {
        "plans": [
          {
            "name": "Local",
            "usd_month": 0,
            "note": "Free to run on your own hardware"
          }
        ],
        "source": "https://ollama.com/library/qwen3-coder",
        "status": "verified",
        "notes": "qwen3-coder:30b is a 19 GB download with a 256K context window."
      },
      "last_checked": "2026-10-21"
    },
    {
      "id": "opencode",
      "name": "OpenCode",
      "url": "https://opencode.ai",
      "kind": "coding-agent",
      "summary": "Open-source, terminal-first coding agent. Bring your own key or run local models.",
      "open_source": true,
      "pricing": {
        "plans": [
          {
            "name": "Open source",
            "usd_month": 0,
            "note": "Pay only for model usage"
          }
        ],
        "source": "https://opencode.ai/docs/providers/",
        "status": "verified",
        "notes": null
      },
      "last_checked": "2026-10-21"
    }
  ],
  "model": "Qwen3-Coder 30B via Ollama (qwen3-coder:30b)",
  "mcp_servers": [],
  "config": {
    "filename": "terminal",
    "language": "bash",
    "code": "ollama pull qwen3-coder:30b\n# one-command launcher from the Ollama library page:\nollama launch opencode --model qwen3-coder"
  },
  "starter_prompt": "You are running fully offline. Read the repository structure first and summarize it in 10 lines. Then <task>. Make small edits, run the tests after each change, and stop to ask if a change touches more than 3 files.",
  "monthly_cost": {
    "min_usd": 0,
    "max_usd": 0,
    "note": "Software is free. You need hardware: the qwen3-coder:30b download is 19 GB, so plan for a machine with enough RAM/VRAM to hold it."
  },
  "steps": [
    "Install Ollama and pull qwen3-coder:30b.",
    "Install OpenCode, or use the ollama launch command.",
    "Run the starter prompt inside your repo.",
    "Keep tasks small; local models do best with tight scopes."
  ],
  "why_this_works": "Nothing is sent to a cloud API, and the 256K context window is large enough for repo-scale work.",
  "watch_out": "Local models are slower and weaker than frontier cloud models on hard tasks. Ollama also lists a cloud variant (480b-cloud); don't use it if privacy is the point.",
  "unverified": [],
  "tested": false,
  "last_checked": "2026-10-21"
}