The name should be general, descriptive of the service being integrated, easy to infer from the task description, and without version numbers. Use it to ground design choices in named patterns, trade-offs and examples.

This document provides guidance on creating comprehensive evaluations for MCP servers. Evaluations test whether LLMs can effectively use your MCP server to answer realistic, complex questions using only the tools provided. Use it to navigate the topic and choose relevant methods, papers or tools.

Create MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. The quality of an MCP server is measured by how well it enables LLMs to accomplish real-world tasks. Use it to make implementation decisions and avoid common dead ends.

List files the agent wrote to /mnt/session/outputs/ during a session, then. Use it to ground design choices in named patterns, trade-offs and examples.

For conceptual overview (tool definitions, tool choice, tips), see shared/tool-use-concepts.md. Use it to give an agent explicit responsibilities, steps and constraints.

For operator instructions that arrive mid-conversation (mode switches, injected state), append {role: "system", ..} to messages instead of editing top-level system — this preserves the cached prefix and carries operator authority. Must follow a user message (or an assista. Use it as a repeatable review, validation or hardening pass.

This file covers the conceptual foundations of tool use with the Claude API. For language-specific code examples, see the python/, typescript/, or other language folders. For decision heuristics on which tools to expose, how to manage context in long-running agents, and cachi. Use it to ground design choices in named patterns, trade-offs and examples.

This file covers how to design prompt-building code for effective caching. For language-specific syntax, see the Prompt Caching section in each language's README or single-file doc. Use it as a repeatable review, validation or hardening pass.

How to move existing code to newer Claude models. Covers breaking changes, deprecated parameters, and drop-in replacements for retired models. Use it as a repeatable review, validation or hardening pass.

Anthropic can POST to your HTTPS endpoint when a Managed Agents resource changes state — an alternative to holding an SSE stream or polling. Payloads are thin (event type + resource IDs only); on receipt, fetch the resource for current state. Every delivery is HMAC-signed. Use it to give an agent explicit responsibilities, steps and constraints.

Override defaults for individual tools. This example enables everything except bash:. Use it to ground design choices in named patterns, trade-offs and examples.

With config.type: "self hosted", the agent loop stays on Anthropic's orchestration layer but tool execution moves to infrastructure you control — bash, file ops, and code run inside your container, so filesystem contents and network egress never leave your environment. Use it as a repeatable review, validation or hardening pass.

A scheduled deployment runs an agent on a recurring cron schedule — each firing creates a session autonomously. Use it for predictable-cadence work: nightly triage, weekly compliance scans, hourly monitors. Use it as a repeatable review, validation or hardening pass.

Managed Agents provisions a container per session as the agent's workspace. The agent loop runs on Anthropic's orchestration layer; the container is where the agent's tools execute — bash commands, file operations, code. You create a persisted Agent config (model, system pr. Use it to ground design choices in named patterns, trade-offs and examples.

An outcome elevates a session from conversation to work : you state what "done" looks like, and the harness runs an iterate → grade → revise loop until the artifact meets the rubric, hits max iterations, or is interrupted. A separate grader (independent context windo. Use it to ground design choices in named patterns, trade-offs and examples.

Claude Managed Agents is a hosted agent: Anthropic runs the agent loop and provisions a sandboxed container per session where the agent's tools execute (or your own worker, with a self hosted environment — see shared/managed-agents-self-hosted-sandboxes.md). You supply an a. Use it to ground design choices in named patterns, trade-offs and examples.

A coordinator agent can delegate to other agents within one session. All agents share the container and filesystem ; each runs in its own thread — a context-isolated event stream with its own conversation history, model, system prompt, tools, MCP servers, and skills (from. Use it to ground design choices in named patterns, trade-offs and examples.

Sessions are ephemeral by default — when one ends, anything the agent learned is gone. A memory store is a workspace-scoped collection of small text documents that persists across sessions. When a store is attached to a session (via resources[]), it is mounted into the cont. Use it as a repeatable review, validation or hardening pass.

Send events to a session via POST /v1/sessions/{id}/events. Use it to ground design choices in named patterns, trade-offs and examples.

Creating a session requires an environment id. Environments are reusable configuration templates for spinning up containers in Anthropic's infrastructure — you might create different environments for different use cases (e.g. data visualization vs web development, with diff. Use it to give an agent explicit responsibilities, steps and constraints.