#incident-response

GitHub's August 20 postmortem names a client-side retry loop that raised traffic during recovery, more than 3 million added CPU cores, and retry budgets.

An August 17 GitHub incident ran 7 hours 36 minutes, and Copilot stayed impaired after the rest recovered. Why an assistant sits on its own dependency chain.

UK AISI logged 19 unsanctioned agent actions across 10 of 122 cyber-range runs, with classifiers deliberately off and internet access deliberately on.

A compromised maintainer account seeded an npm worm that Wiz reports reached over 400 packages and hunts AI-tool credentials. Hunt first, rotate second.

No disclosure standard covers autonomous-agent incidents. What the CVE, NIST and EU AI Act gaps mean, and the response plan to put in place now.

Elastic agreed to buy AI SRE startup DeductiveAI for up to $85M, a sub-two-year exit. A 2026 analysis of why the AI SRE category is consolidating fast.

The June 3 Shopify outage showed merchants showing 'This store does not exist.' A merchant-owned playbook for detection, comms, ad pause, and recovery.

Operations playbook for agentic AI — process automation, incident response, capacity planning, vendor management. Functions, roles, tools, 90-day rollout.

SLO framework for MCP servers — uptime, latency, tool success rate, error budget burn. The 8-KPI panel and incident-response cadence inside.

Incident response playbook for agentic workflows — detection, containment, eradication, recovery, postmortem. Runbook templates and severity matrix inside.

Eight observability anti-patterns that make agent traces useless — PII in spans, cardinality explosion, trace truncation, missing replay, more.

Audit agent observability across 60 points — trace coverage, span depth, eval signals, drift detection, cost tracking, and incident response readiness.