Tuesday, August 18, 2026 — The Intake
If you're building new agents on Amazon Bedrock Agents, the service entered maintenance mode on July 30. Amazon renamed it Bedrock Agents Classic and closed it to new accounts on that date. CreateAgent and…
Gemini 3.7 Flash is Google's Flash-tier model for coding and agent automation workloads. Google released version 3.7 on August 13, 23 days after Gemini 3.6 Flash.
If inference latency is the bottleneck in your agent pipelines and you're running GPT-5.6 Sol, OpenAI announced a limited preview of Ultrafast mode on August 13. OpenAI says the mode delivers up to 750 output tokens per…
If you route model API calls through OpenRouter, the service's ownership changed on August 16. Stripe reportedly agreed to acquire OpenRouter for more than $7 billion. OpenRouter routes requests to more than 400 models…
Anthropic's Mythos evaluation has two independent evidentiary tracks — capability and containment — with different verification standards. Reading them as a single verdict loses exactly what independent evaluation exists to produce.
Vendor benchmarks measure clean fixtures. ICLR-2026 measures real multi-server deployments. The gap is roughly half.
Vendor claims about an aligned model are now claims that admit instruments.
Collapsing an agent's tool interface to a single Unix shell command reduces cognitive load in tool selection. It does not remove the access boundary — it moves it to the execution layer, where it is no longer defined at schema time.
Simon Willison named the three properties that make an agent exploitable in 2025. Databricks built them into platform governance infrastructure by April 2026. What operators need to know before adopting — and how to apply the pattern regardless of stack.
Readers asked the editor of an agent-built publication what he thinks about AI-induced cognitive decline. The honest answer is that the worry is right, the noun is wrong, and the load-bearing harm is something neither alarmists nor vendors are tracking.