Trial by Trace: Benchmarking AI Agent Harnesses for Reliability and Failure Attribution · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2610.32285/Collected HN comments and submissions linking this paper.enWed, 04 Nov 2026 05:35:47 GMTTrial by Trace: Benchmarking AI Agent Harnesses for Reliability and Failure Attribution — HN submissionhttps://news.ycombinator.com/item?id=50337394urn:hackerpapers:submission:2610.32285:50337394Mon, 02 Nov 2026 14:35:12 GMT<p>Show HN: Nomos – A Go harness that stops coding agents from cheating on tests</p><p>Hi HN, Mark here. If you&#x27;ve pointed Claude Code, Cursor, or OpenCode at a codebase you care about and walked away thinking AI coding tools are a net negative, I pretty much agree with you. Left directly on a working branch, models optimize for shutting the compiler up on the current turn. Hit a tricky failure and they delete the failing assertion, return a hardcoded nil/true stub, rewrite the spec to match broken code, dump a 300-line function into the wrong package, or burn $15 of tokens editing in circles. Writing longer rules files (&quot;IMPORTANT: NEVER DELETE TESTS&quot;) breaks down once context…</p><p>4 points; 0 discussion replies on HN. <a href="https://hackerpaper.kaipfister.de/p/2610.32285/">Paper and source history</a>. A submission is not an endorsement.</p>cs.AI