Daily digest · 2026-10-29 (UTC)
What broke for AI agents on 29 Oct 2026
9 incidents on 2 status pages; 1 probe finding from our own round-trips.
Final. Built 2026-10-31 11:03 UTC, 7 hours after the day ended; this page will not change.
What broke
- Cloudflare: “Increased Durable Objects, D1 and Workflows Error Rates”minor impact · opened 14:16 UTC · resolved 16:33 UTC (2 h 17 min) · Artifacts | D1 | Durable Objects | Workers Assets | Workflowsstatus page ↗provider-incidents v77
- Cloudflare: “Increased errors and latency for CNI 2.0 traffic in LAX”minor impact · opened 06:55 UTC · resolved 08:28 UTC (93 minutes) · Cloud Network Interconnect (CNI) | Los Angeles, United States - (LAX)status page ↗provider-incidents v76
- Cloudflare: “Increased Errors for Durable Objects and Downstream Services in Eastern North America (ENAM)”minor impact · opened 16:36 UTC · resolved 16:50 UTC (50 minutes) · Artifacts | D1 | Durable Objects | R2status page ↗provider-incidents v77
- Cloudflare: “Increased HTTP 429 Rate Limit Errors on Cloudflare Dashboard”minor impact · opened 10:31 UTC · resolved 12:10 UTC (99 minutes) · Dashboard | Support Sitestatus page ↗provider-incidents v76
- Cloudflare: “Increased Worker services errors”minor impact · opened 09:22 UTC · resolved 11:49 UTC (2 h 26 min) · Containers | D1 | Workers Assets | Workflowsstatus page ↗provider-incidents v76
- Cloudflare: “Issues with Durable Objects availability in Eastern Europe”minor impact · opened 10:06 UTC · resolved 11:47 UTC (101 minutes) · Durable Objectsstatus page ↗provider-incidents v76
- Cloudflare: “Network Performance Issues in Queretaro, MX”minor impact · opened 01:50 UTC · resolved 00:35 UTC (70 minutes) · Queretaro, Mexico - (QRO)status page ↗provider-incidents v74
- Cloudflare: “R2 Elevated Latency in Eastern Europe”minor impact · opened 19:57 UTC · resolved 21:50 UTC (113 minutes) · R2status page ↗provider-incidents v78
- GitHub: “Disruption with Copilot Cloud services”minor impact · opened 10:19 UTC · resolved 12:49 UTC (2 h 30 min)status page ↗provider-incidents v76
- mcp-registry (registry.modelcontextprotocol.io): 2 of 8 round-trips failedno answer ×2 · first at 18:57 UTC · measured from macbook-kragent-api-observatory v120
Shipped
No new major version of the SDKs and tools we follow.
14 other releases on the day
- @google/genai 2.32.0 (npm)released 06:40 UTC · previous 2.31.0agent-sdk-releases v90
- @langchain/core 1.3.2 (npm)released 17:55 UTC · previous 1.3.1agent-sdk-releases v92
- ai 7.0.157 (npm)released 17:10 UTC · previous 7.0.156agent-sdk-releases v91
- crewAIInc/crewAI 1.15.33published 07:36 UTC · previous 1.15.32release notes ↗agent-tool-releases v90
- google-genai 2.34.0 (pypi)released 06:43 UTC · previous 2.33.0agent-sdk-releases v90
- langchain-ai/langchain langchain-core==1.7.2published 19:04 UTC · previous langchain-core==1.7.1release notes ↗agent-tool-releases v92
- langchain-ai/langchain langchain-core==1.7.3published 20:56 UTC · previous langchain-core==1.7.2release notes ↗agent-tool-releases v92
- langchain-core 1.7.3 (pypi)released 20:55 UTC · previous 1.7.1agent-sdk-releases v92
- vercel/ai @ai-sdk/[email protected]published 17:14 UTC · previous @ai-sdk/[email protected]release notes ↗agent-tool-releases v91
- vercel/ai @ai-sdk/[email protected]published 17:06 UTC · previous @ai-sdk/[email protected]release notes ↗agent-tool-releases v91
- vercel/ai @ai-sdk/[email protected]published 17:14 UTC · previous @ai-sdk/[email protected]release notes ↗agent-tool-releases v91
- vercel/ai [email protected]published 17:03 UTC · previous [email protected]release notes ↗agent-tool-releases v91
- vercel/ai [email protected]published 18:47 UTC · previous [email protected]release notes ↗agent-tool-releases v92
- vercel/ai [email protected]published 17:10 UTC · previous [email protected]release notes ↗agent-tool-releases v91
Measured on the day
- 184 round-trips to 23 endpoints agents depend on, 2 failedmodel APIs, registries, chains, the x402 facilitator, status pages and two baselines · measured from macbook-kragent-api-observatory v120
- 462 of 500 remote MCP endpoints probed answered a handshake (92.4%), on 449 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 v28
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