AI Weekly
Every Monday, the 6 stories that mattered most from the past seven days of the brief: why each one made the cut, with a link to the full analysis.
- Oct 19–25 2026 · 6 stories This week the risk moved from model quality to the plumbing around it: API versions, gateways and dependencies
- Claude API ships 2026-10-21, its first new version since 2023-06-01: temperature, top_p and top_k now return errors, and new major SDKs move to Apache 2.0
- LiteLLM 1.107.0 was a poisoned release: malicious code ran through Python startup hooks and stole credentials without LiteLLM ever being imported
- Anthropic donates the Agent Skills standard to the Linux Foundation's AAIF: more than 70 agent platforms, about 1 million skills published and nearly 280 million installs
- OpenAI adds 12-hour prompt caching to the API, with writes at 2.5x the input price
- GitHub package firewall enters public preview, screening npm traffic between coding agents, CI and the registry
- OpenAI's frontier models and Codex workflows are coming to Google Cloud through Model Garden and Google Cloud Marketplace