Anthropic Managed Agents adds fleet orchestration, Outcomes, and self-improving Dreaming
Anthropic ships fleet orchestration, Outcomes goal-spec, and Dreaming for Managed Agents. Dreaming lets agents self-improve across sessions without fine-tuning.
Anthropic ships fleet orchestration, Outcomes goal-spec, and Dreaming for Managed Agents. Dreaming lets agents self-improve across sessions without fine-tuning.
Anthropic ships Claude Code Routines — a (prompt, repo, connectors) config that runs on Anthropic infra, triggered by schedule, HTTP POST, or GitHub event.