modus-agendi/managed-agent-control-mcp

**Start, observe, and interact with Claude Managed Agents from any MCP client** (Claude.ai, Claude Code, Cursor, mcp-remote, your own agent, …).
Claude Managed Agents run on Anthropic's platform — each is an agent definition
(model, system prompt, tools, skills) that executes inside a sandbox environment.
This MCP server puts a **remote control** on those agents: connect it to **any MCP
client** and you can launch an agent, watch what it does, reply to it, approve the
tools it wants to run, and stop it. Claude.ai is the showcase client (do it all
from a normal conversation), but the same tools work from Claude Code, Cursor,
mcp-remote, the MCP Inspector, or a custom MCP client. It is the runtime
companion to the Terraform provider
that *defines* agents declaratively.
It runs as a local **stdio** server, a generic **HTTP container**, or on **AWS
Lambda**, with pluggable inbound auth (static **bearer**, generic **OIDC**, or an
AWS **Cognito** preset).
> [!NOTE]
> This is a community project. It is not maintained by, endorsed by, or affiliated
> with Anthropic. The Managed Agents API is a beta API (managed-agents-2026-04-01).
How it works
You drive a loop: **discover** an agent → **start** a session → **observe** by
polling events → **interact** (reply / interrupt / approve tools) → **end**.
Because MCP tool calls are request/response, observation is by polling
(session_events / session_get) — there is no live stream into the chat.
Quickstart (local, <2 min)
You need an ANTHROPIC_API_KEY with Managed Agents access and
uv.
Explore the tools interactively with the MCP Inspector:
Register it with **Claude Code** (a project-scoped .mcp.json is included):
Once published to PyPI, you can skip the clone: uvx managed-agent-control-mcp.
Tools
| Tier | Tool | Does |
|------|------|------|
| Discover | agent_list / agent_get | Find an agent and inspect its config |
| | environment_list / environment_get | Find a sandbox environment |
| | vault_list / vault_get | Find a credential vault to attach at session_start |
| | memory_store_list / memory_store_get | Find a persistent memory store |
| Start | session_start | Create a session and (optionally) send the first instruction |
| Observe | session_get | Status (idle/running/…) + token usage |
| | session_list | List sessions |
| | session_events | Poll the agent's output/activity (cursor + type filter) |
| Interact | session_message | Send a message / continue a turn |
| | session_interrupt | Stop or redirect a running agent |
| | session_respond | Approve/deny a tool the agent is waiting on |
| Destructive 🔒 | session_archive / session_delete | Archive (keep history) or delete |
See docs/tools.md for every argument and return shape.
Configuration
Only ANTHROPIC_API_KEY is required for local use. Full reference in
docs/configuration.md.
| Env var | Purpose |
|---|---|
| ANTHROPIC_API_KEY | **Required.** Operator key the server acts with. |
| ANTHROPIC_BASE_URL | Override the API base URL (gateways/testing). |
| MCP_AUTH_MODE | Inbound auth for HTTP: bearer, oidc, cognito (comma-separated to combine). |