Hacker NewsNobody Gets Promoted for Simplicity 833 Points HNTerrible Software·high signalXBlueskyLinkedInCopy linkResonant essay on organizational incentives rewarding complexity. HN connects to AI code proliferation risks. 833 points, 473 comments.SourceSource pageTerrible Software↳ Follow the threadPolicy dependency / Stack layerA Cheap Read-Only Verifier Captures Nearly All the False-Pass Benefit of a Full Planning StackarXiv 2609.20474Stack layer / Threat patternKilocode 7.7.5 adds @model, @past-chats and @worktrees mentions and fixes a session-switching memory leakGitHubPolicy dependency / Stack layerClaude Code 2.1.277 reads AGENTS.md when there is no CLAUDE.md, and subagent output now carries an anti-spoofing headerClaude Code changelogStack layer / Update threadAnthropic publishes an R&D Automation Index: Claude 'leads' 26% of its own AI research work as of August 2026Anthropic Institute (corroborated by The Washington Post and Euronews)Stack layer / ContrastAn Auto-Optimized Agent Harness Can Cheat the Whole Benchmark, Not Just the TasksarXiv 2609.18366Policy dependency / Stack layer43 Machine-Checkable EU AI Act Criteria That Run in CI and Emit Article-Indexed EvidencearXiv 2609.20016Stack layer / ContrastRecording an Agent Run at Its Non-Deterministic Boundaries Turns an Incident Into a CI Regression TestarXiv 2609.20625Stack layer / ContrastHarness-Layer Auto-Research Cut Agent Token Traffic 44.7-49.0% at Equal Task PerformancearXiv 2609.20519