Claude Managed Agents: Anthropic's Hosted Runtime for Enterprise AI Agents (2026)
Claude Managed Agents announcement and launch details: Anthropic's hosted API runtime handles tool use, persistent memory, sub-agent orchestration, and observability so enterprise teams ship production agents without managing the runtime themselves.
Anthropic announced Claude Managed Agents, a hosted runtime that handles tool use, memory, retrieval, and orchestration for production agents. The platform targets teams who want Claude's agent capabilities without operating the underlying glue.
What's included
- Hosted tool execution — declare tools as functions, Anthropic handles the runtime and retries.
- Persistent memory — conversation and task state stored across turns with TTL controls.
- Workflow primitives — sub-agent dispatch, fan-out/fan-in, and long-running task suspension.
- Observability — per-step traces, latency breakdown, and cost attribution per agent invocation.
Why it matters
The bottleneck for enterprise agent deployment has been the operational surface, not the model. By moving the orchestration layer behind a managed API, Anthropic is betting that the long tail of teams adopting agents would rather buy a runtime than build one.
For TheRouter customers: managed agents are accessible through the standard
anthropic/ model namespace via TheRouter's gateway. No code changes
required for existing agent workloads — the routing layer transparently
proxies the new endpoints.
Claude Managed Agents Gets Scheduled Deployments, Vault Env Vars, and session_thread_id
Anthropic’s Claude Managed Agents update adds scheduled deployments, Vault environment-variable credentials, and session_thread_id webhooks so teams can replace external cron jobs, stop passing secrets through prompts, and route multi-agent events with clearer correlation.
Claude Managed Agents Gets Self-Hosted Sandboxes and MCP Tunnels: What the Split-Plane Architecture Means for Operators
Anthropic's Managed Agents now lets teams run tool execution in their own infrastructure while keeping orchestration on Anthropic's side. Here's what that split-plane model means for routing, compliance, and private MCP access.
Anthropic Model Hardware Standard: The New Safety Boundary for Physical AI Agents
Anthropic Model Hardware Standard turns lab devices into discoverable agent tools. For operators, the critical work is routing authority, safety limits, and audit paths before code touches hardware.