Daily digest · 2026-11-09 (UTC)

What broke for AI agents on 9 Nov 2026

9 incidents on 3 status pages.

Final. Built 2026-11-10 07:40 UTC, 7 hours after the day ended; this page will not change.

What broke

1. GitHub: “Incident with Actions and Checks API”critical impact · opened 18:12 UTC · resolved 18:59 UTC (47 minutes) · Actions | Pages | Webhooksstatus page ↗provider-incidents v106
2. GitHub: “Degraded availability for GPT-5.3 Codex and GPT-5.6 models in Copilot”major impact · opened 23:16 UTC · resolved Nov 10 00:05 UTC (49 minutes) · Copilot | Copilot AI Model Providersstatus page ↗provider-incidents v106
3. GitHub: “Incident with Webhooks and Actions”major impact · opened 10:00 UTC · resolved 11:50 UTC (110 minutes) · Actions | Webhooksstatus page ↗provider-incidents v104
4. Cloudflare: “Fedramp Customers have Delayed Purges”minor impact · opened Nov 6 17:47 UTC · identified as of Nov 10 00:00 UTC · CDN Cache Purgestatus page ↗provider-incidents v97
5. Cloudflare: “Changes made to Workers may be delayed or disrupted”minor impact · opened 16:15 UTC · resolved Nov 10 01:46 UTC (9 h 31 min) · Workersstatus page ↗provider-incidents v107
6. Cloudflare: “Network Performance Issues in Eastern North America”minor impact · opened 16:34 UTC · resolved 20:12 UTC (3 h 38 min) · Networkstatus page ↗provider-incidents v106
7. Cloudflare: “Spectrum Availability Issue in Phoenix”minor impact · opened Nov 7 00:59 UTC · resolved 14:36 UTC (3 days) · Spectrumstatus page ↗provider-incidents v105
8. GitHub: “Increased timeouts due to surge in anonymous users”minor impact · opened 05:59 UTC · resolved 09:09 UTC (3 h 11 min)status page ↗provider-incidents v104
9. OpenAI: “Elevated error rates with some U.S. business Mastercards”the feed gives no impact level · first seen by our reader Nov 7 00:19 UTC (the feed gives no start time) · monitoring as of Nov 10 00:00 UTCstatus page ↗provider-incidents v99

Shipped

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

25 other releases on the day

1. @anthropic-ai/claude-agent-sdk 0.3.321 (npm)released 19:03 UTC · previous 0.3.319agent-sdk-releases v124
2. @anthropic-ai/sdk 1.7.0 (npm)released 18:56 UTC · previous 1.6.0agent-sdk-releases v124
3. @langchain/core 1.3.6 (npm)released 23:57 UTC · previous 1.3.5agent-sdk-releases v124
4. ai 7.0.166 (npm)released 02:55 UTC · previous 7.0.165agent-sdk-releases v121
5. anthropic 2.7.0 (pypi)released 18:52 UTC · previous 2.6.0agent-sdk-releases v124
6. anthropics/anthropic-sdk-python v2.7.0published 18:52 UTC · previous v2.6.0release notes ↗agent-tool-releases v122
7. anthropics/anthropic-sdk-typescript sdk-v1.7.0published 18:52 UTC · previous sdk-v1.6.0release notes ↗agent-tool-releases v122
8. anthropics/claude-agent-sdk-python v0.2.180published 20:09 UTC · previous v0.2.179release notes ↗agent-tool-releases v122
9. anthropics/claude-code v2.1.321published 19:22 UTC · previous v2.1.319release notes ↗agent-tool-releases v122
10. BerriAI/litellm v1.108.4published 23:00 UTC · previous v1.108.3release notes ↗agent-tool-releases v123
11. claude-agent-sdk 0.2.180 (pypi)released 20:07 UTC · previous 0.2.179agent-sdk-releases v124
12. crewAIInc/crewAI 1.15.39published 01:07 UTC · previous 1.15.38release notes ↗agent-tool-releases v121
13. langchain-ai/langchain langchain-fireworks==1.7.3published 20:54 UTC · previous langchain-fireworks==1.7.2release notes ↗agent-tool-releases v123
14. litellm 1.108.4 (pypi)released 22:56 UTC · previous 1.108.3agent-sdk-releases v124
15. ollama/ollama v0.40.16published 18:21 UTC · previous v0.40.15release notes ↗agent-tool-releases v123
16. openai 3.48.0 (pypi)released 16:13 UTC · previous 3.46.0agent-sdk-releases v123
17. openai 7.54.0 (npm)released 16:15 UTC · previous 7.53.0agent-sdk-releases v123
18. openai/openai-node v7.54.0published 16:04 UTC · previous v7.53.0release notes ↗agent-tool-releases v122
19. openai/openai-python v3.48.0published 16:12 UTC · previous v3.47.0release notes ↗agent-tool-releases v122
20. vercel/ai @ai-sdk/[email protected]published 02:56 UTC · previous @ai-sdk/[email protected]release notes ↗agent-tool-releases v121
21. vercel/ai @ai-sdk/[email protected]published 02:56 UTC · previous @ai-sdk/[email protected]release notes ↗agent-tool-releases v121
22. vercel/ai @ai-sdk/[email protected]published 02:56 UTC · previous @ai-sdk/[email protected]release notes ↗agent-tool-releases v121
23. vercel/ai [email protected]published 19:46 UTC · previous [email protected]release notes ↗agent-tool-releases v122
24. vercel/ai [email protected]published 02:57 UTC · previous [email protected]release notes ↗agent-tool-releases v121
25. vercel/ai [email protected]published 02:54 UTC · previous [email protected]release notes ↗agent-tool-releases v121

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 v164
2. 385 of 800 remote MCP endpoints probed answered a handshake (48.1%), on 565 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 v39

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

2026-11-122026-11-112026-11-102026-11-092026-11-082026-11-072026-11-062026-11-052026-11-042026-11-032026-11-022026-11-012026-10-312026-10-302026-10-292026-10-282026-10-272026-10-262026-10-252026-10-24