Daily digest · 2026-10-30 (UTC)
What broke for AI agents on 30 Oct 2026
6 incidents on 3 status pages.
Final. Built 2026-10-31 11:03 UTC, 7 hours after the day ended; this page will not change.
What broke
- GitHub: “Disruption with GitHub services”critical impact · opened 14:31 UTC · resolved 15:02 UTC (32 minutes) · Issues | Pagesstatus page ↗provider-incidents v80
- GitHub: “Incident with Actions, Pages, and Webhooks”major impact · opened 17:50 UTC · resolved 19:23 UTC (93 minutes) · Actions | Pages | Webhooksstatus page ↗provider-incidents v80
- Cloudflare: “Increased Errors for Browser Run”minor impact · opened 02:41 UTC · resolved 03:48 UTC (67 minutes) · Browser Runstatus page ↗provider-incidents v79
- GitHub: “Incident with Dependabot”minor impact · opened 20:59 UTC · resolved 23:29 UTC (2 h 30 min)status page ↗provider-incidents v81
- OpenAI: “Elevated errors for GPT Image 2 API image edits”the feed gives no impact level · the feed gives no start time · resolved 18:36 UTC · Imagesstatus page ↗provider-incidents v80
- OpenAI: “Elevated errors with search in Dots, ChatGPT Work, and certain ChatGPT thinking experiences”the feed gives no impact level · first seen by our reader 20:06 UTC (the feed gives no start time) · resolved 21:14 UTC · ChatGPT Work | dotsstatus page ↗provider-incidents v81
Shipped
No new major version of the SDKs and tools we follow.
17 other releases on the day
- @anthropic-ai/claude-agent-sdk 0.3.314 (npm)released 00:14 UTC · previous 0.3.312agent-sdk-releases v93
- ai 7.0.158 (npm)released 15:38 UTC · previous 7.0.157agent-sdk-releases v94
- ai 7.0.159 (npm)released 22:18 UTC · previous 7.0.158agent-sdk-releases v95
- anthropics/claude-agent-sdk-python v0.2.175published 01:50 UTC · previous v0.2.174release notes ↗agent-tool-releases v93
- anthropics/claude-code v2.1.314published 01:06 UTC · previous v2.1.312release notes ↗agent-tool-releases v93
- claude-agent-sdk 0.2.175 (pypi)released 01:48 UTC · previous 0.2.174agent-sdk-releases v93
- langchain-ai/langgraph 1.2.19published 22:23 UTC · previous 1.2.18release notes ↗agent-tool-releases v95
- langchain-ai/langgraph checkpointsqlite==3.1.3published 22:18 UTC · previous checkpointsqlite==3.1.2release notes ↗agent-tool-releases v95
- langchain-ai/langgraph sdk==0.4.11published 22:17 UTC · previous sdk==0.4.10release notes ↗agent-tool-releases v95
- vercel/ai @ai-sdk/[email protected]published 15:43 UTC · previous @ai-sdk/[email protected]release notes ↗agent-tool-releases v94
- vercel/ai @ai-sdk/[email protected]published 15:43 UTC · previous @ai-sdk/[email protected]release notes ↗agent-tool-releases v94
- vercel/ai @ai-sdk/[email protected]published 15:43 UTC · previous @ai-sdk/[email protected]release notes ↗agent-tool-releases v94
- vercel/ai @ai-sdk/[email protected]published 15:43 UTC · previous @ai-sdk/[email protected]release notes ↗agent-tool-releases v94
- vercel/ai [email protected]published 19:33 UTC · previous [email protected]release notes ↗agent-tool-releases v95
- vercel/ai [email protected]published 19:29 UTC · previous [email protected]release notes ↗agent-tool-releases v95
- vercel/ai [email protected]published 15:39 UTC · previous [email protected]release notes ↗agent-tool-releases v94
- vercel/ai [email protected]published 22:18 UTC · previous [email protected]release notes ↗agent-tool-releases v95
Measured 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 v124
- 397 of 500 remote MCP endpoints probed answered a handshake (79.4%), on 480 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 v29
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
2026-11-012026-10-312026-10-302026-10-292026-10-282026-10-272026-10-262026-10-252026-10-24