Show HN: Mingbird – an agent harness that makes a 2B model finish real tasks

Show HN: Mingbird – an agent harness that makes a 2B model finish real tasks 中文版 · Apache-2.0 · Windows 10/11 · Linux & macOS (experimental) · Offline-first Same 2B model: 0.017 → 0.821. You've been told local models need a big discrete GPU.

This PolicyChange is relevant to the technology intelligence record because it involves GitHub, OpenAI, Llama, Qwen. The source article should remain the factual reference for follow-up coverage.

• 中文版 · Apache-2.0 · Windows 10/11 · Linux & macOS (experimental) · Offline-first Same 2B model: 0.017 → 0.821.
• You've been told local models need a big discrete GPU.
• On an ordinary laptop — integrated graphics, 16–32 GB of RAM — the same 2–9B models that stall in cloud-style frameworks deliver finished artifacts here, because the failures you have seen are harness defects, not model defects.
• Measured end to end; all 288 cells public.
• The quality floor is held up by a 461-test regression suite — 435 unit tests plus 26 integration tests, the prefill zero-growth assertion among them.
• The suite contains no live-Ollama end-to-end test : nothing in it needs a running backend or a network.