{"date":"2026-11-04","title":"What happened in AI — Nov 4, 2026","generated_at":"2026-11-04T22:30:00Z","intro":["Heavier agent scaffolding lost to simpler setups: a 100-task benchmark found Claude Code with more thinking beat a 37-skill, 17-agent harness that cost 8× more.","NVIDIA's result that 16K contexts with compaction hold accuracy points the same way: cut context and orchestration, not capability. Kimi K3.1 Pro's day-0 vLLM support shows open-weight serving catching up fast."],"highlights":["Extended thinking beat a 37-skill, 17-agent harness that cost 8× more.","NVIDIA: frontier agents lose little accuracy at 16K context with compaction; long context costs up to 7× more per token.","Cleric's ~40-agent rewrite made evals 5× faster and 100× cheaper.","Kimi K3.1 Pro (2.8T MoE, open weights) had day-0 vLLM support.","Glow Labs: coding agents are installing deprecated MCP servers with known flaws.","H200 loses money fully busy; B300 profits at 35% utilization."],"article_count":32,"categories":[{"name":"Agent harnesses: less scaffolding, durable memory","slug":"agent-harnesses-memory","summary":"A 100-task benchmark found extended thinking beat elaborate harnesses, while three new tools move agent memory into portable stores.","articles":[{"title":"We Tested Coding Agent Harnesses. Thinking Harder Beat All of Them","summary":"Across 100 tasks, a 37-skill, 17-agent harness cost 8× more and scored below vanilla Claude Code with more thinking.","source":"hackernews_ai","url":"https://surgehq.ai/blog/benchmarking-coding-agent-harnesses","published":"Wed, 04 Nov 2026 20:52:13 +0000"},{"title":"Renor: one memory for all your AI agents","summary":"Apache-2.0 Markdown-notes memory recalled over MCP, with clients for Claude Code, Codex, Cursor, Gemini and VS Code.","source":"hackernews_ai","url":"https://github.com/renorai/renor-engine","published":"Wed, 04 Nov 2026 17:06:10 +0000"},{"title":"Engraphis: local-first, bi-temporal memory for coding agents","summary":"Inspectable, self-hostable memory with code-aware recall and bi-temporal history, served over MCP.","source":"hackernews_ai","url":"https://github.com/Coding-Dev-Tools/engraphis","published":"Wed, 04 Nov 2026 04:52:44 +0000"},{"title":"Mentat: one memory for all my AI tools","summary":"Linked notes in a private git repo, served to Claude and OpenCode from a small Cloudflare Worker.","source":"hackernews_ai","url":"https://blog.benjamin.ie/blog/mentat","published":"Wed, 04 Nov 2026 21:50:55 +0000"},{"title":"DeepSeek Open-Sources Super-Agent, a Coding Agent That Works From Team Chat","summary":"DeepSeek released a coding agent driven from team chat, as an open-source project.","source":"search_cn_open_weight_labs","publisher_name":"Technology Org","publisher_domain":"technology.org","url":"https://news.google.com/rss/articles/CBMioAFBVV95cUxPOWRQWmRtMFZFaDVDWGpNYXg2WV9pcVlKcVItSXlrS0twTU55X3VPZU1OSFFOUGhQUmlMcXg5dkVIemtCMFYydFdxc2lDOTQzUFZYNjhZOEpBTHlzdmt2V0JRRXlqRG9DZ0xnSFI3Y0stYnNyRFZpNzg5ekh6czVoVW5vUk40WEE5ZVA1ZVNseXplX3Z3b2hGTUFiN1RvS3Vk?oc=5","published":"Wed, 04 Nov 2026 09:57:00 GMT"}]},{"name":"Evals and context: cheaper runs, shorter windows","slug":"evals-context-cost","summary":"Two results cut the cost of agent operations: Cleric's 40-agent eval rewrite and NVIDIA's finding that 16K contexts with compaction hold accuracy.","articles":[{"title":"How we made evals 100× cheaper with multi-agent architecture","summary":"Cleric split one investigation agent into ~40 loosely coupled agents, running evals 5× faster for cents.","source":"hackernews_ai","url":"https://cleric.ai/blog/how-we-made-evals-100x-cheaper-with-multi-agent-architecture","published":"Wed, 04 Nov 2026 16:17:30 +0000"},{"title":"Nvidia Study: AI Agents Rarely Need Long Context Windows","summary":"Claude Opus 4.5 and GPT 5.6 Sol lose little accuracy at 16K with frequent compaction; long-context serving costs up to 7× more per token.","source":"hackernews_ai","url":"https://i10x.ai/news/nvidia-long-context-compaction-agents-study","published":"Wed, 04 Nov 2026 14:17:09 +0000"},{"title":"Evalship: find the flaws in your LLM evals","summary":"Watches every pull request and auto-optimizes prompts against your existing eval suite.","source":"hackernews_ai","url":"https://evalship.com/","published":"Wed, 04 Nov 2026 10:03:08 +0000"}]},{"name":"Inference economics and open-weight serving","slug":"inference-open-weight-serving","summary":"Kimi K3.1 Pro had day-0 serving support, while two measurements show gateway language and GPU choice matter more than engine choice.","articles":[{"title":"vLLM Delivers Day-0 Support for Kimi K3.1 Pro","summary":"Moonshot's 2.8T-parameter (104B active) open-weight MoE runs on NVIDIA, AMD, TPU and Alibaba PPU from launch.","source":"vllm_blog","url":"https://vllm.ai/blog/2026-11-04-kimi-k31-pro","published":"Wed, 04 Nov 2026 00:00:00 GMT"},{"title":"Verda Inference, in Preview with Kimi K3.1 Pro","summary":"Verda adds managed model-as-a-service inference for open-weight frontier models, launching with Kimi K3.1 Pro.","source":"hackernews_ai","url":"https://verda.com/blog/verda-inference-in-preview","published":"Wed, 04 Nov 2026 16:58:53 +0000"},{"title":"Is hosting an open-source inference engine profitable?","summary":"H200 loses money even fully busy; B300 turns a profit at 35% utilization, so the GPU matters more than the engine.","source":"hackernews_ai","url":"https://userecurrence.ai/research/is-hosting-open-source-inference-profitable","published":"Wed, 04 Nov 2026 02:58:34 +0000"},{"title":"Go vs. Python for an LLM Gateway: What We Measured","summary":"Reproducible benchmark of KTMS (Go) against LiteLLM Proxy and eight other gateways on latency, TTFT and memory.","source":"hackernews_ai","url":"https://blog.ai-process.dev/2026/10/26/go-vs-python-llm-gateway/","published":"Wed, 04 Nov 2026 18:06:08 +0000"},{"title":"How big is the open model economy today?","summary":"Interconnects uses its open-model dashboard to show adoption metrics are noisy and need an informed read.","source":"interconnects","url":"https://www.interconnects.ai/p/how-well-can-we-measure-open-model","published":"Wed, 04 Nov 2026 16:35:44 GMT"}]},{"name":"Agent security and identity","slug":"agent-security-identity","summary":"Coding agents are installing retired MCP servers on their own, and NIST and Cloudflare each publish controls for identifying and governing agents.","articles":[{"title":"AI coding agents are installing abandoned MCP servers with known security issues","summary":"Glow Labs' \"MCPhantom\" pattern: agents pick deprecated MCP servers that no longer receive security fixes.","source":"hackernews_ai","url":"https://www.glow.io/blogs/mcphantom-ai-coding-agents-are-installing-vulnerable-deprecated-mcp-servers-from-training-data","published":"Wed, 04 Nov 2026 10:58:43 +0000"},{"title":"Design Tenets for Agent Identity Management Systems","summary":"NIST's second post on agent identity, authentication and authorization sets design tenets for identity systems.","source":"hackernews_ai","url":"https://www.nist.gov/blogs/cybersecurity-insights/design-tenets-agent-identity-management-systems","published":"Wed, 04 Nov 2026 02:42:44 +0000"},{"title":"See the agents, set the terms: new visibility and control for personal agents","summary":"Cloudflare adds visibility into which personal agents visit a site and lets owners set terms for them.","source":"cloudflare_blog","url":"https://blog.cloudflare.com/personal-agents-visibility-control/","published":"Wed, 04 Nov 2026 14:00:00 GMT"},{"title":"Daybreak Red is now available in AWS GovCloud (US-East)","summary":"GPT-5.6 Cyber reaches approved customers through Amazon Bedrock for authorized cybersecurity work.","source":"openai_blog","url":"https://openai.com/index/daybreak-red-aws-govcloud","published":"Wed, 04 Nov 2026 07:00:00 GMT"}]},{"name":"Distribution and spend: Qwen on iPhone, Claude Code economics","slug":"distribution-spend","summary":"Distribution and spend drive platform strategy: Apple's Qwen trial in China, OpenAI buying into Codex tooling, and Anthropic's Claude Code margins.","articles":[{"title":"Apple Begins China Trial Of AI Tools Powered By Alibaba's Qwen","summary":"Apple's China beta links Qwen with Siri, Writing Tools and Visual Intelligence.","source":"search_cn_open_weight_labs","publisher_name":"Dataconomy","publisher_domain":"dataconomy.com","url":"https://news.google.com/rss/articles/CBMilgFBVV95cUxQR3lraDN3cHQ5OTRPSHlHWWJEN2lFZ01lbWxXLXBJeDZlblpxcjN6YWVtaGw3TE44Zy03TkVHNkJjOW5pdmJIdS1wQ1pxTGx3bDZIOVZKVWJ2NFdtQnM3OGRjejVWT3lSUVR2TnpnN1c4RzhTQ0h4ak1mS3M2VFVhQ0lMc3ZFN1BiVmhVZFZmMk1KQVRLcWc?oc=5","published":"Wed, 04 Nov 2026 10:22:33 GMT"},{"title":"OpenAI to acquire Pierre Computer Company, Inc.","summary":"The acquisition expands Codex's tools for building, understanding and reviewing code.","source":"openai_blog","url":"https://openai.com/index/openai-to-acquire-pierre","published":"Wed, 04 Nov 2026 14:30:00 GMT"},{"title":"Anthropic launched tools to show companies what their AI spending is buying","summary":"Claude Code reportedly hit an operating profit in Q2 with 900% net revenue retention.","source":"hackernews_ai","url":"https://techcrunch.com/2026/10/28/anthropic-just-launched-tools-to-show-companies-what-their-ai-spending-is-buying/","published":"Wed, 04 Nov 2026 04:26:22 +0000"}]}],"video":{"youtube_id":"KgLSjr8ynnM","title":"LLM Digest Daily Explainer: Nov 4, 2026","duration_s":112.6}}