Daily digest · 2026-11-14 (UTC)

What broke for AI agents on 14 Nov 2026

10 incidents on 3 status pages.

Final. Built 2026-11-15 07:11 UTC, 7 hours after the day ended; this page will not change.

What broke

  1. Anthropic: “Some users unable to sign up with Apple on the Claude iOS app”major impact · opened 07:11 UTC · investigating as of Nov 15 00:00 UTC · claude.aistatus page ↗provider-incidents v124
  2. Anthropic: “Elevated purchasing errors on claude.ai”major impact · opened 02:54 UTC · resolved 04:23 UTC (89 minutes) · claude.aistatus page ↗provider-incidents v123
  3. Cloudflare: “Cloudflare is currently having an issue with the r8.im registry used for Replicate model publishing and builds is currently restricted”minor impact · opened Nov 13 14:55 UTC · resolved 00:51 UTC (9 h 56 min) · Replicatestatus page ↗provider-incidents v123
  4. Cloudflare: “Cloudflare is seeing an issue with Workers API availability”minor impact · opened Nov 13 15:31 UTC · resolved 13:37 UTC (22 h 6 min) · Workersstatus page ↗provider-incidents v125
  5. Cloudflare: “Elevated error rates on R2 buckets in West North America”minor impact · opened 20:18 UTC · resolved 20:20 UTC (40 minutes) · R2status page ↗provider-incidents v126
  6. Cloudflare: “Increased errors for R2 and Tiered Cache”minor impact · opened 05:30 UTC · resolved 06:45 UTC (75 minutes) · Chicago, United States - (ORD) | Newark, United States - (EWR) | R2status page ↗provider-incidents v124
  7. Cloudflare: “Issues with 1.1.1.1 public resolver in Canada”minor impact · opened Nov 13 23:53 UTC · resolved 01:01 UTC (68 minutes) · Recursive DNSstatus page ↗provider-incidents v123
  8. OpenAI: “Delayed Costs Data in the Compliance API”the feed gives no impact level · first seen by our reader Nov 13 18:20 UTC (the feed gives no start time) · monitoring as of Nov 15 00:00 UTC · Compliance APIstatus page ↗provider-incidents v122
  9. OpenAI: “Elevated errors affecting the Microsoft 365 connector”the feed gives no impact level · first seen by our reader Nov 13 00:20 UTC (the feed gives no start time) · resolved 00:18 UTC · Connectors/Appsstatus page ↗provider-incidents v122
  10. OpenAI: “Elevated errors with ChatGPT billing and Ads billing”the feed gives no impact level · the feed gives no start time · resolved 05:27 UTCstatus page ↗provider-incidents v123

Shipped

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

6 other releases on the day
  1. BerriAI/litellm v1.106.10published 00:34 UTC · previous v1.108.8release notes ↗agent-tool-releases v136
  2. BerriAI/litellm v1.108.9published 22:00 UTC · previous v1.106.10release notes ↗agent-tool-releases v138
  3. langchain-ai/langgraph 1.2.22published 01:43 UTC · previous 1.2.21release notes ↗agent-tool-releases v137
  4. langchain-ai/langgraph checkpoint==4.3.2published 01:31 UTC · previous checkpoint==4.3.1release notes ↗agent-tool-releases v137
  5. langchain-ai/langgraph checkpointpostgres==3.2.1published 01:39 UTC · previous checkpointpostgres==3.2.0release notes ↗agent-tool-releases v137
  6. langchain-ai/langgraph prebuilt==1.1.2published 01:33 UTC · previous prebuilt==1.1.1release notes ↗agent-tool-releases v137

Measured on the day

  1. 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 v184
  2. 477 of 800 remote MCP endpoints probed answered a handshake (59.6%), on 553 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 v44

How this is made

  • 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.

Earlier days

2026-11-212026-11-202026-11-192026-11-182026-11-172026-11-162026-11-152026-11-142026-11-132026-11-122026-11-112026-11-102026-11-092026-11-082026-11-072026-11-062026-11-052026-11-042026-11-032026-11-022026-11-012026-10-312026-10-302026-10-292026-10-282026-10-272026-10-262026-10-252026-10-24

Terms change on 2026-11-30 — what changes2026-11-30 시행 약관 변경 안내