{"playbook": "production-ops", "title": "Running AI in production", "description": "Monitoring, tracing and alerting for LLM apps, incident response, rollouts, model upgrades and provider migrations, reliability.", "url": "https://wpnews.pro/playbooks/production-ops", "count": 3, "recommendations": [{"advice": "Run autonomous real-time resource allocation for agent workloads on Kubernetes, starting in read-only mode on your hardest environments, and reclaim idle agent memory via swap to avoid provisioning for peak spikes.", "takeaway": "Agents that idle and then burst hold memory the whole time and crash 80% more often than microservices, so ScaleOps reallocates idle agent resources to other workloads and uses memory swap to provision memory only on demand, which it says cuts cloud costs 70% to 80% and makes agents 90% more cost-ef", "article_url": "https://wpnews.pro/news/idle-then-spiking-why-ai-agents-are-breaking-kubernetes-resource-planning", "title": "Idle, then spiking: Why AI agents are breaking Kubernetes resource planning", "source_url": "https://siliconangle.com/2026/11/17/autonomous-orchestration-helps-kubernetes-keep-up-ai-agents-kubeconna/", "published_at": "2026-11-17T20:35:39+00:00"}, {"advice": "Audit shared networking gateways and failover paths for latent defects, and keep spare capacity above the 14 per cent shortfall that worsened Microsoft's outage before traffic shifts cascade.", "takeaway": "Microsoft's outage stemmed from a networking gateway software defect introduced in July that, combined with a Teams failover and roughly 14 per cent of server capacity already unavailable, left over 80 per cent of gateway servers unresponsive — showing Copilot outages can trace to shared infrastruct", "article_url": "https://wpnews.pro/news/did-ai-cause-the-microsoft-outage-or-us-flight-groundings-what-officials-have", "title": "Did AI Cause the Microsoft Outage or US Flight Groundings? What Officials Have Revealed About Both Incidents", "source_url": "https://www.ibtimes.co.uk/microsoft-outage-faa-ground-stop-not-caused-ai-1832941", "published_at": "2026-11-17T19:54:03+00:00"}, {"advice": "Benchmark your own repo's wall-clock and runner minutes before adopting an agent-optimized CI harness; StarSling's 9-repo test averaged 10x speedup and 70% less runner time, but per-repo results ranged from 1.3x to 43x.", "takeaway": "Across 9 open source repos, StarSling's agent-driven CI harness averaged a 10x wall-clock speedup with 70% less runner time, from 2.2x on prompts.chat to 36x on mirotalksfu.", "article_url": "https://wpnews.pro/news/announcing-the-ci-challenge", "title": "Announcing the CI Challenge", "source_url": "https://starsling.dev/blog/ci-challenge", "published_at": "2026-11-17T00:00:00+00:00"}]}