|
The Daily Reconcile.
All the cloud native news that’s fit to reconcile
| Tuesday, October 13, 2026 |
Two pages |
Front page · tap to read the PDF
Platform engineering
Spotify's merged code changes went from about 8,100 a month to roughly 17,000 in the year to August 2026, across nearly 3,000 production services. Fleet Management has merged over 2.5 million automated pull requests, and a background agent runs Claude in Kubernetes pods. Then a scheduling bug cut ingestion throughput by 10 per cent, and an automated dependency upgrade passed every check and failed in production. The remedies were platform work: monitoring, rollback capacity, and changes scheduled in the owning team's hours. Today's tip: kubectl diff -f pod.json — see what an apply would change first.
What the extra changes were | Quality work, 2025 | | 27 | | Quality work, 2026 | | 31 | | Maintenance, 2025 | | 31 | | Maintenance, 2026 | | 25 |
Bars to one scale, highest value 31 |
Also in today’s two pages
| → | vLLM lets decode hand the KV cache back to prefill: GLM-5-FP8 time to first token falls from 2,334 ms to 702 ms at 20k tokens | | → | CVE-2026-97229, CVSS 10.0: a CI/CD regex compiler gives unauthenticated code execution in GitLab, patched in 19.4.2, 19.3.4 and 19.2.8 | | → | A fake AWS Terraform provider plans your infrastructure and drops a Rust backdoor | | → | Dapr 1.18.6 stops spoofed caller identity headers reaching target apps | | → | Krateo PlatformOps: the portal stopped fetching 15.3 MB a call to keep 49 KB |
|
Want it on paper?
The PDF is laid out for exactly this: A4, two pages, printed on one sheet. It is how this paper is read every morning.
A4 · double-sided · flip on the long edge
|
Researched, written and laid out by Claude Code. Every item links to its source; always check the original.
Made by Diego Braga, who decides what it covers.
Sources today: InfoQ, vLLM, CNCF, securityonline.info, Zscaler ThreatLabz, Argo CD, Dapr, KubeVirt, Cluster API and GitHub
You get this because you asked for it, and for nothing else. We keep your address and your paper preference, and no other data.
Privacy ·
Source on GitHub
|