Daily digest · 2026-11-07 (UTC)

What broke for AI agents on 7 Nov 2026

5 incidents on 3 status pages.

In progress. As of 05:05 UTC, rebuilt every hour; it freezes at 07:00 UTC on 8 Nov 2026.

What broke

Cloudflare: “Fedramp Customers have Delayed Purges”minor impact · opened Nov 6 17:47 UTC · identified as of Nov 8 00:00 UTC · CDN Cache Purgestatus page ↗provider-incidents v97
Cloudflare: “Spectrum Availability Issue in Phoenix”minor impact · opened 00:59 UTC · monitoring as of Nov 8 00:00 UTC · Spectrumstatus page ↗provider-incidents v99
Cloudflare: “Network Performance Issues in Chicago”minor impact · opened 05:34 UTC · resolved 05:12 UTC (17 minutes) · Chicago, United States - (ORD) | Networkstatus page ↗provider-incidents v99
npm: “Minor issues impacting package publish”minor impact · opened 14:28 UTC · resolved 15:05 UTC (37 minutes) · Package publishingstatus page ↗provider-incidents v100
OpenAI: “Elevated error rates with some U.S. business Mastercards”the feed gives no impact level · first seen by our reader 00:19 UTC (the feed gives no start time) · monitoring as of Nov 8 00:00 UTCstatus page ↗provider-incidents v99

No new major version of the SDKs and tools we follow.

9 other releases on the day

184 round-trips to 23 endpoints agents depend on, 0 failedmodel APIs, registries, chains, the x402 facilitator, status pages and two baselines · measured from macbook-kragent-api-observatory v156
485 of 788 remote MCP endpoints probed answered a handshake (61.5%), on 559 hostsa different slice of the official registry each day, so the share is not compared across days · one attempt per endpoint, an auth challenge counts as an answer · measured from macbook-krmcp-server-liveness v37

Every line comes from one of WITAN's public datasets and links the version its records landed in: provider-incidents, agent-api-observatory, mcp-server-liveness, agent-sdk-releases, agent-tool-releases. They are free to read.
Incidents are what the providers' own status pages published (Anthropic, OpenAI, GitHub, Cloudflare, npm, Vercel), read every 6 hours; scheduled maintenance is left out. Follow a line's status page link for the provider's text.
Round-trips and MCP handshakes are WITAN's own measurements from one vantage point, named on each line. A failure there can be the network between us and the service; it is not a verdict on the service. A latency line needs the day's median at least 2× the previous 7 days' and 250 ms more.
Releases are those of the SDKs and agent tools the datasets follow; pre-releases are left out.
Days are UTC. A day is rebuilt every hour while it runs and frozen 7 hours after it ends. Model prices are not here: that dataset's collector is off. WITAN is a testnet preview run by one developer.