https://codorc.com/en/Coding news, useful research, and a little orcish curiosity. Follow the changes that matter to the work you do.enhttps://codorc.com/en/news/gemini-decisions-api-calibration/https://codorc.com/en/news/gemini-decisions-api-calibration/Wed, 11 Nov 2026 21:52:14 GMTGoogle introduced the Gemini Decisions API on November 11 for bounded multimodal choices. The announcement describes scoring candidate options rather than generating a free-form answer.https://codorc.com/en/news/control-room-agent-operations/https://codorc.com/en/news/control-room-agent-operations/Wed, 11 Nov 2026 21:44:00 GMTOpenAI described Control Room on November 9 as a system for tracking and stopping model workloads across its research infrastructure. The useful design lesson is operational visibility.https://codorc.com/en/news/claude-fable-mythos-5-2/https://codorc.com/en/news/claude-fable-mythos-5-2/Wed, 11 Nov 2026 21:44:00 GMTAnthropic announced Fable 5.2 and Mythos 5.2 on November 9. The company distinguishes broadly available Fable from verification-gated Mythos access.https://codorc.com/en/news/images-2-6-production-workflow/https://codorc.com/en/news/images-2-6-production-workflow/Wed, 11 Nov 2026 21:44:00 GMTOpenAI introduced Images 2.6 on November 4, describing image generation, editing and GIF capabilities. Product teams still need an asset pipeline.https://codorc.com/en/news/scaling-git-for-agents/https://codorc.com/en/news/scaling-git-for-agents/Wed, 11 Nov 2026 21:44:00 GMTOpenAI published its account of Pando, a Git-compatible backend, on November 2. Increased agent activity turns repository infrastructure into a capacity question.https://codorc.com/en/news/ai-training-crawl-preferences/https://codorc.com/en/news/ai-training-crawl-preferences/Wed, 11 Nov 2026 21:44:00 GMTOpenAI announced support for a shared robots.txt training-preference signal on October 30. Publishers should distinguish discovery from model training.https://codorc.com/en/news/github-secret-detection/https://codorc.com/en/news/github-secret-detection/Wed, 11 Nov 2026 21:44:00 GMTGitHub described a context-aware secret-detection model on October 13. Its report focuses on detection quality and scanning efficiency.https://codorc.com/en/news/claude-haiku-5-5-routing/https://codorc.com/en/news/claude-haiku-5-5-routing/Wed, 11 Nov 2026 21:44:00 GMTAnthropic introduced Claude Haiku 5.5 on October 7 for high-volume work including summarization, subagents and browser use. Routing remains an engineering decision.https://codorc.com/en/news/reviewbench-code-review/https://codorc.com/en/news/reviewbench-code-review/Wed, 11 Nov 2026 21:44:00 GMTGitHub introduced ReviewBench on October 5 as an open benchmark for AI code review. The important question is whether review catches consequential defects.https://codorc.com/en/news/september-ai-developer-roundup/https://codorc.com/en/news/september-ai-developer-roundup/Wed, 11 Nov 2026 21:44:00 GMTGoogle’s October 2 recap collected its September AI announcements, including Gemini 4 Argon. A monthly recap is a useful discovery map, not an adoption checklist.https://codorc.com/en/news/gpt-6-1-sol-evaluation/https://codorc.com/en/news/gpt-6-1-sol-evaluation/Wed, 11 Nov 2026 21:44:00 GMTOpenAI introduced GPT-6.1 Sol for coding, computer use and professional tasks on September 29. The release makes model selection a workload question.https://codorc.com/en/news/prompt-caching-economics/https://codorc.com/en/news/prompt-caching-economics/Wed, 11 Nov 2026 21:44:00 GMTOpenAI announced expanded GPT-6 prompt-caching controls and diagnostics on September 22. Repeated context deserves its own performance budget.https://codorc.com/en/news/claude-opus-5-5-release/https://codorc.com/en/news/claude-opus-5-5-release/Wed, 11 Nov 2026 21:44:00 GMTAnthropic released Claude Opus 5.5 on September 22, positioning it for agentic coding and knowledge work. Its claims need workload-specific evaluation.https://codorc.com/en/news/rag-mcp-skills-different-jobs/https://codorc.com/en/news/rag-mcp-skills-different-jobs/Wed, 11 Nov 2026 21:44:00 GMTGitHub’s September 18 podcast article explored debates about reading code, retrieval and agent tools. The useful follow-up is to separate the layers in an agent system.https://codorc.com/en/news/agent-consistency-repeated-runs/https://codorc.com/en/news/agent-consistency-repeated-runs/Wed, 11 Nov 2026 21:44:00 GMTIBM Research published work on repeated-run agent consistency on September 15. Its central distinction is between average success and succeeding every time the same task is repeated.