Doc2Harness: New benchmark finds best coding agent passes only 53% across 329 cases
_akhaliq · x · 2026-10-16
Doc2Harness, from ICT-CAS, is billed as the first benchmark for long-horizon harness engineering of coding agents. It evaluates 19 frontier models across 328 test cases, with the best model reaching only a 53% pass rate — even the strongest systems fail roughly half the time, underscoring how far agents remain from reliable long-horizon engineering work.
More from coding & agent
- Vercel Goes Full Backend: Python GA, Containers, and Daemons — cramforce · 2026-10-16
- WorldMemArena Accepted at NeurIPS: Memory Storage ≠ Memory Use — xwang_lk · 2026-10-16
- Vercel Ship Demos Container Deploys for Go, Ruby, Java and More — edwin · 2026-10-16
- TRL adds day-0 support for fine-tuning Muse Spark Open's 816B MoE with LoRA on 64 H100s — QGallouedec · 2026-10-16
- Databricks engineers land impactful PRs on day 2 with AI dev tools, vs week 3 before — Yuchenj_UW · 2026-10-16
- Perplexity open-sources SPACE, the sandbox platform powering its Computer agents — AravSrinivas · 2026-10-16