Recommended reading: Ruoshi's blog argues the model is not dumb, the harness is misconfigured

推荐阅读若石写的这篇博客：模型不是笨，是 Harness 没配好

Ruoshi’s blog attributes multi-step agent failures to harness design, not model ability, and lays out four engineering rules plus a one-day minimum setup. The post cites failures after context exceeds 70%, log compression from 32K to 7K tokens, external state in state.json, schema validation, and local retries; the post does not disclose quantified success-rate gains. What matters for practitioners is execution constraints, externalized state, and independent evaluation rather than more prompt tuning.

Why it matters: HKR-H lands on the contrarian hook: agent failure is blamed on harness design, not model IQ. HKR-K and HKR-R land via concrete knobs—70% context threshold, 32K→7K logs, external state, schema retry—but this is still a reposted recommendation with no disclosed win-rate lift.