Anthropic's Claude Platform is now generally available on AWS, offering full Claude API feature parity with AWS IAM authentication, CloudTrail audit logging, and a single AWS invoice that retires against existing commitments.
Anthropic launched Dreaming—a scheduled process that reviews agent session history and updates memory between tasks—as a research preview in Claude Managed Agents at Code with Claude 2026. Outcomes, multiagent orchestration, and webhooks also moved to public beta.
HN latched onto the money leak before the bug itself. A report that Claude Managed Agents append a malware reminder to every file read, then sometimes refuse to edit code anyway, turned into a broader argument about opaque token spend and whether agent harnesses deserve more scrutiny.
Claude said on April 8, 2026 that Managed Agents lets teams define tasks, tools, and guardrails while Anthropic runs the agent infrastructure. Anthropic's official materials describe a composable API suite for cloud-hosted, versioned agents, with advanced capabilities like outcomes, memory, and multi-agent orchestration in limited research preview.
On April 8, 2026, Anthropic highlighted a new engineering post describing Managed Agents, its hosted service for long-running agent work on the Claude Platform. Anthropic says the system separates session, harness, and sandbox layers so agents can recover more cleanly from failure and connect to customer infrastructure with fewer assumptions.