Tools

Tools extend what agents can do—letting them fetch real-time data, execute code, query external databases, and take actions in the world.

Under the hood, tools are callable functions with well-defined inputs and outputs that get passed to a chat model. The model decides when to invoke a tool based on the conversation context, and what input arguments to provide.

Create tools

Basic tool definition

The simplest way to create a tool is by importing the tool function from the langchain package. You can use zod to define the tool's input schema:

import * as z from "zod"
import { tool } from "langchain"

const searchDatabase = tool(
  ({ query, limit }) => `Found ${limit} results for '${query}'`,
  {
    name: "search_database",
    description: "Search the customer database for records matching the query.",
    schema: z.object({
      query: z.string().describe("Search terms to look for"),
      limit: z.number().describe("Maximum number of results to return"),
    }),
  }
);

Access context

Tools are most powerful when they can access runtime information like conversation history, user data, and persistent memory. This section covers how to access and update this information from within your tools.

Context

Context provides immutable configuration data that is passed at invocation time. Use it for user IDs, session details, or application-specific settings that shouldn't change during a conversation.

Tools can access an agent's runtime context through the config parameter. Configure a checkpointer and pass context with a stable thread_id so the conversation persists across invocations:

import * as z from "zod";
import { uuid7 } from "langsmith";
import { ChatOpenAI } from "@langchain/openai";
import { MemorySaver } from "@langchain/langgraph";
import { createAgent, tool } from "langchain";

const getUserName = tool(
  (_, config) => {
    return config.context.user_name;
  },
  {
    name: "get_user_name",
    description: "Get the user's name.",
    schema: z.object({}),
  },
);

const contextSchema = z.object({
  user_name: z.string(),
});

const agent = createAgent({
  model: new ChatOpenAI({ model: "gpt-5.5" }),
  tools: [getUserName],
  checkpointer: new MemorySaver(),
  contextSchema,
});

const threadId = uuid7();
const threadConfig = {
  configurable: { thread_id: threadId },
  context: { user_name: "John Smith" },
};

let result = await agent.invoke(
  {
    messages: [{ role: "user", content: "What is my name?" }],
  },
  threadConfig,
);
console.log(result.messages.at(-1)?.content);

result = await agent.invoke(
  {
    messages: [{ role: "user", content: "What was my name again?" }],
  },
  threadConfig,
);
console.log(result.messages.at(-1)?.content);

Long-term memory (Store)

The BaseStore provides persistent storage that survives across conversations. Unlike state (short-term memory), data saved to the store remains available in future sessions.

Access the store through config.store. The store uses a namespace/key pattern to organize data:

import * as z from "zod";
import { createAgent, tool } from "langchain";
import { InMemoryStore } from "@langchain/langgraph";
import { ChatOpenAI } from "@langchain/openai";

const store = new InMemoryStore();

// Access memory
const getUserInfo = tool(
  async ({ user_id }) => {
    const value = await store.get(["users"], user_id);
    console.log("get_user_info", user_id, value);
    return value;
  },
  {
    name: "get_user_info",
    description: "Look up user info.",
    schema: z.object({
      user_id: z.string(),
    }),
  }
);

// Update memory
const saveUserInfo = tool(
  async ({ user_id, name, age, email }) => {
    console.log("save_user_info", user_id, name, age, email);
    await store.put(["users"], user_id, { name, age, email });
    return "Successfully saved user info.";
  },
  {
    name: "save_user_info",
    description: "Save user info.",
    schema: z.object({
      user_id: z.string(),
      name: z.string(),
      age: z.number(),
      email: z.string(),
    }),
  }
);

const agent = createAgent({
  model: new ChatOpenAI({ model: "gpt-5.5" }),
  tools: [getUserInfo, saveUserInfo],
  store,
});

// First session: save user info
await agent.invoke({
  messages: [
    {
      role: "user",
      content: "Save the following user: userid: abc123, name: Foo, age: 25, email: foo@langchain.dev",
    },
  ],
});

// Second session: get user info
const result = await agent.invoke({
  messages: [
    { role: "user", content: "Get user info for user with id 'abc123'" },
  ],
});

console.log(result);
// Here is the user info for user with ID "abc123":
// - Name: Foo
// - Age: 25
// - Email: foo@langchain.dev

Stream writer

Stream real-time updates from tools during execution. This is useful for providing progress feedback to users during long-running operations.

Use config.writer to emit custom updates:

import * as z from "zod";
import { tool, ToolRuntime } from "langchain";

const getWeather = tool(
  ({ city }, config: ToolRuntime) => {
    const writer = config.writer;

    // Stream custom updates as the tool executes
    if (writer) {
      writer(`Looking up data for city: ${city}`);
      writer(`Acquired data for city: ${city}`);
    }

    return `It's always sunny in ${city}!`;
  },
  {
    name: "get_weather",
    description: "Get weather for a given city.",
    schema: z.object({
      city: z.string(),
    }),
  }
);

Execution info

Access thread ID, run ID, and retry state from within a tool via runtime.execution_info:

import { tool } from "langchain";
import * as z from "zod";

const logExecutionContext = tool(
  async (_input, runtime) => {
    const info = runtime.executionInfo;
    console.log(`Thread: ${info.threadId}, Run: ${info.runId}`);  // [!code highlight]
    console.log(`Attempt: ${info.nodeAttempt}`);
    return "done";
  },
  {
    name: "log_execution_context",
    description: "Log execution identity information.",
    schema: z.object({}),
  }
);

Server info

When your tool runs on LangGraph Server, access the assistant ID, graph ID, and authenticated user via runtime.server_info:

import { tool } from "langchain";
import * as z from "zod";

const getAssistantScopedData = tool(
  async (_input, runtime) => {
    const server = runtime.serverInfo;
    if (server != null) {
      console.log(`Assistant: ${server.assistantId}, Graph: ${server.graphId}`);  // [!code highlight]
      if (server.user != null) {
        console.log(`User: ${server.user.identity}`);  // [!code highlight]
      }
    }
    return "done";
  },
  {
    name: "get_assistant_scoped_data",
    description: "Fetch data scoped to the current assistant.",
    schema: z.object({}),
  }
);

serverInfo is null when the tool is not running on LangGraph Server.

Tool execution

In LangChain, tools are used by agents (for example via create_agent) and tool error handling is configured through middleware.

For LangGraph workflows, tool execution is handled by ToolNode. See ToolNode for Graph API usage, including how tools can access the current graph state and run-scoped context.

Tool return values

You can choose different return values for your tools:

• Return a string for human-readable results.
• Return an object for structured results the model should parse.
• Return a Command with optional message when you need to write to state.

Return a string

Return a string when the tool should provide plain text for the model to read and use in its next response.

import { tool } from "langchain";
import * as z from "zod";

const getWeather = tool(({ city }) => `It is currently sunny in ${city}.`, {
  name: "get_weather",
  description: "Get weather for a city.",
  schema: z.object({ city: z.string() }),
});

Behavior:

• The return value is converted to a ToolMessage.
• The model sees that text and decides what to do next.
• No agent state fields are changed unless the model or another tool does so later.

Use this when the result is naturally human-readable text.

Return an object

Return an object (for example, a dict) when your tool produces structured data that the model should inspect.

import { tool } from "langchain";
import * as z from "zod";

const getWeatherData = tool(
  ({ city }) => ({
    city,
    temperature_c: 22,
    conditions: "sunny",
  }),
  {
    name: "get_weather_data",
    description: "Get structured weather data for a city.",
    schema: z.object({ city: z.string() }),
  },
);

Behavior:

• The object is serialized and sent back as tool output.
• The model can read specific fields and reason over them.
• Like string returns, this does not directly update graph state.

Use this when downstream reasoning benefits from explicit fields instead of free-form text.

Return multimodal content

Tools are not limited to plain text. When the model supports multimodal tool results, the tool can return standard content blocks so the model receives text, images, and other media in one tool result.

import { tool } from "langchain";
import { z } from "zod";

const captureScreenshot = tool(
  async () => [
    { type: "text", text: "Screenshot of the current page:" },
    { type: "image", url: "https://example.com/page.png" },
  ],
  {
    name: "capture_screenshot",
    description: "Capture a screenshot of the current page.",
    schema: z.object({}),
  }
);

Behavior:

• The return value is converted to a ToolMessage with multimodal content.
• Use message.content_blocks to read the normalized block list after the tool runs.
• The model must support the modalities you return. Check your model's capabilities before returning images, audio, or video.

For block types and provider-specific requirements, see Multimodal messages.

Return a Command

Return a Command when the tool needs to update graph state (for example, setting user preferences or app state). When the Command targets the current graph, include a ToolMessage in the update whose tool call ID matches the current tool call. Every tool call in the message history must have a corresponding ToolMessage.

Use runtime.toolCallId for the tool_call_id parameter.

import { tool, ToolMessage, type ToolRuntime } from "langchain";
import { Command } from "@langchain/langgraph";
import * as z from "zod";

const setLanguage = tool(
  async ({ language }, config: ToolRuntime) => {
    return new Command({
      update: {
        preferredLanguage: language,
        messages: [
          new ToolMessage({
            content: `Language set to ${language}.`,
            tool_call_id: config.toolCallId,
          }),
        ],
      },
    });
  },
  {
    name: "set_language",
    description: "Set the preferred response language.",
    schema: z.object({ language: z.string() }),
  },
);

Behavior:

• The command updates state using update.
• Updated state is available to subsequent steps in the same run.
• Use reducers for fields that may be updated by parallel tool calls.

Use this when the tool is not just returning data, but also mutating agent state.

Return directly from a tool

Set return direct on a tool to short-circuit the agent loop: the agent returns the tool's output to the caller immediately, without sending it back through the model for further processing.

import { ChatOpenAI } from "@langchain/openai";
import { createAgent, tool } from "langchain";
import * as z from "zod";

const fetchOrderStatus = tool(
  ({ order_id }) => {
    return `Order ${order_id} is shipped and will arrive in 2 days.`;
  },
  {
    name: "fetch_order_status",
    description: "Fetch the current status of a customer order.",
    schema: z.object({ order_id: z.string() }),
    returnDirect: true,
  },
);

const agent = createAgent({
  model: new ChatOpenAI({ model: "gpt-4o-mini" }),
  tools: [fetchOrderStatus],
});

const result = await agent.invoke({
  messages: [
    { role: "user", content: "What is the status of order #12345?" },
  ],
});
// The agent returns the tool output directly without another LLM call:
// "Order 12345 is shipped and will arrive in 2 days."

Behavior:

• The tool executes normally and its output is wrapped in a ToolMessage.
• The agent stops looping and returns the tool's output as the final response, bypassing any additional model call.
• Multiple parallel tool calls: When the model calls several tools in one step, all of them execute first. After all tools finish, the agent routes to END only if every tool in that batch has return_direct=True. The final response includes the ToolMessage outputs of every tool that was called in that step.

Use this when:

• The tool's output is the complete, user-ready answer (for example, a lookup that returns a ready-to-display result).
• You want to avoid an extra model call when no additional reasoning is needed.
• You need deterministic, unmodified output: the model cannot rephrase, summarize, or act on the tool result.

Error handling

Handle tool errors using LangChain agent middleware to retry failed tool calls or return custom error messages:

import { createAgent, createMiddleware, ToolMessage } from "langchain";

const handleToolErrors = createMiddleware({
  name: "HandleToolErrors",
  wrapToolCall: async (request, handler) => {
    try {
      return await handler(request);
    } catch (error) {
      return new ToolMessage({
        content: `Tool error: Please check your input and try again. (${error})`,
        tool_call_id: request.toolCall.id!,
      });
    }
  },
});

const agent = createAgent({
  model: "google:gemini-3.6-flash",
  tools: [],
  middleware: [handleToolErrors],
});

import { createAgent, createMiddleware, ToolMessage } from "langchain";

const handleToolErrors = createMiddleware({
  name: "HandleToolErrors",
  wrapToolCall: async (request, handler) => {
    try {
      return await handler(request);
    } catch (error) {
      return new ToolMessage({
        content: `Tool error: Please check your input and try again. (${error})`,
        tool_call_id: request.toolCall.id!,
      });
    }
  },
});

const agent = createAgent({
  model: "openai:gpt-5.5",
  tools: [],
  middleware: [handleToolErrors],
});

import { createAgent, createMiddleware, ToolMessage } from "langchain";

const handleToolErrors = createMiddleware({
  name: "HandleToolErrors",
  wrapToolCall: async (request, handler) => {
    try {
      return await handler(request);
    } catch (error) {
      return new ToolMessage({
        content: `Tool error: Please check your input and try again. (${error})`,
        tool_call_id: request.toolCall.id!,
      });
    }
  },
});

const agent = createAgent({
  model: "anthropic:claude-sonnet-5",
  tools: [],
  middleware: [handleToolErrors],
});

import { createAgent, createMiddleware, ToolMessage } from "langchain";

const handleToolErrors = createMiddleware({
  name: "HandleToolErrors",
  wrapToolCall: async (request, handler) => {
    try {
      return await handler(request);
    } catch (error) {
      return new ToolMessage({
        content: `Tool error: Please check your input and try again. (${error})`,
        tool_call_id: request.toolCall.id!,
      });
    }
  },
});

const agent = createAgent({
  model: "openrouter:z-ai/glm-5.2",
  tools: [],
  middleware: [handleToolErrors],
});

import { createAgent, createMiddleware, ToolMessage } from "langchain";

const handleToolErrors = createMiddleware({
  name: "HandleToolErrors",
  wrapToolCall: async (request, handler) => {
    try {
      return await handler(request);
    } catch (error) {
      return new ToolMessage({
        content: `Tool error: Please check your input and try again. (${error})`,
        tool_call_id: request.toolCall.id!,
      });
    }
  },
});

const agent = createAgent({
  model: "fireworks:accounts/fireworks/models/glm-5p2",
  tools: [],
  middleware: [handleToolErrors],
});

import { createAgent, createMiddleware, ToolMessage } from "langchain";

const handleToolErrors = createMiddleware({
  name: "HandleToolErrors",
  wrapToolCall: async (request, handler) => {
    try {
      return await handler(request);
    } catch (error) {
      return new ToolMessage({
        content: `Tool error: Please check your input and try again. (${error})`,
        tool_call_id: request.toolCall.id!,
      });
    }
  },
});

const agent = createAgent({
  model: "baseten:zai-org/GLM-5.2",
  tools: [],
  middleware: [handleToolErrors],
});

import { createAgent, createMiddleware, ToolMessage } from "langchain";

const handleToolErrors = createMiddleware({
  name: "HandleToolErrors",
  wrapToolCall: async (request, handler) => {
    try {
      return await handler(request);
    } catch (error) {
      return new ToolMessage({
        content: `Tool error: Please check your input and try again. (${error})`,
        tool_call_id: request.toolCall.id!,
      });
    }
  },
});

const agent = createAgent({
  model: "ollama:north-mini-code-1.0",
  tools: [],
  middleware: [handleToolErrors],
});

State injection

Tools access graph state through ToolRuntime. See Access context for state, context, store, and streaming APIs.

For more details on accessing state, context, and long-term memory from tools, see Access context.

Dynamic tool selection

With dynamic tools, the set of tools available to the agent is modified at runtime rather than defined all upfront. Not every tool is appropriate for every situation. Too many tools may overwhelm the model (overload context) and increase errors; too few limit capabilities. Dynamic tool selection enables adapting the available toolset based on authentication state, user permissions, feature flags, or conversation stage.

There are two approaches depending on whether tools are known ahead of time:

import { createMiddleware, tool } from "langchain";
import { createDeepAgent } from "deepagents";

const stateBasedTools = createMiddleware({
    name: "StateBasedTools",
    wrapModelCall: (request, handler) => {
        // Read from State: check authentication and conversation length
        const state = request.state as typeof request.state & {
            authenticated?: boolean;
        };
        const isAuthenticated = state.authenticated ?? false;
        const messageCount = state.messages.length;

        let filteredTools = request.tools;

        // Only enable sensitive tools after authentication
        if (!isAuthenticated) {
            filteredTools = request.tools.filter(
                (t: any) => typeof t.name === "string" && t.name.startsWith("public_"),
            );
        } else if (messageCount < 5) {
            filteredTools = request.tools.filter(
                (t: any) => typeof t.name === "string" && t.name !== "advanced_search",
            );
        }

        return handler({ ...request, tools: filteredTools });
    },
});

const agent = await createDeepAgent({
    model: "claude-sonnet-4-6",
    tools: tools,
    middleware: [stateBasedTools] as any,
});

import { createMiddleware } from "langchain";
import { createDeepAgent, StoreBackend } from "deepagents";
import * as z from "zod";
import { InMemoryStore } from "@langchain/langgraph";

const contextSchema = z.object({
  userId: z.string(),
});

const storeBasedTools = createMiddleware({
  name: "StoreBasedTools",
  contextSchema,
  wrapModelCall: async (request, handler) => {
    const userId =
      (request.runtime?.context as { userId?: string } | undefined)?.userId ??
        "user-123";

    // Read from Store: get user's enabled features
    const runtimeStore = request.runtime?.store as InMemoryStore | undefined;
    const rawFlags = (await runtimeStore?.get(
      ["features"],
      userId as string,
    )) as unknown;
    const featureFlags = rawFlags as FeatureFlags | undefined;

    let filteredTools = request.tools;

    if (featureFlags) {
      const enabledFeatures = featureFlags.enabledTools || [];
      filteredTools = request.tools.filter((t) =>
        enabledFeatures.includes(t.name as string)
      );
    }

    return handler({ ...request, tools: filteredTools });
  },
});

const agent = await createDeepAgent({
  model: "claude-sonnet-4-6",
  backend: new StoreBackend(),
  store,
  checkpointer,
  tools,
  middleware: [storeBasedTools] as any,
});

import * as z from "zod";
import { createMiddleware } from "langchain";
import { createDeepAgent } from "deepagents";

const contextSchema = z.object({
  userRole: z.string(),
});

const contextBasedTools = createMiddleware({
  name: "ContextBasedTools",
  contextSchema,
  wrapModelCall: (request, handler) => {
    // Read from Runtime Context: get user role
    const userRole = request.runtime.context.userRole;

    let filteredTools = request.tools;

    if (userRole === "admin") {
      // Admins get all tools
    } else if (userRole === "editor") {
      filteredTools = request.tools.filter((t) => t.name !== "delete_data");
    } else {
      filteredTools = request.tools.filter(
        (t) => (t.name as string).startsWith("read_"),
      );
    }

    return handler({ ...request, tools: filteredTools });
  },
});

const agent = await createDeepAgent({
  model: "claude-sonnet-4-6",
  store,
  checkpointer,
  tools,
  middleware: [contextBasedTools] as any,
});

import { createAgent, createMiddleware, tool } from "langchain";
import * as z from "zod";

// A tool that will be added dynamically at runtime
const calculateTip = tool(
  ({ billAmount, tipPercentage = 20 }) => {
    const tip = billAmount * (tipPercentage / 100);
    return `Tip: $${tip.toFixed(2)}, Total: $${(billAmount + tip).toFixed(2)}`;
  },
  {
    name: "calculate_tip",
    description: "Calculate the tip amount for a bill",
    schema: z.object({
      billAmount: z.number().describe("The bill amount"),
      tipPercentage: z.number().default(20).describe("Tip percentage"),
    }),
  }
);

const dynamicToolMiddleware = createMiddleware({
  name: "DynamicToolMiddleware",
  wrapModelCall: (request, handler) => {
    // Add dynamic tool to the request
    // This could be loaded from an MCP server, database, etc.
    return handler({
      ...request,
      tools: [...request.tools, calculateTip],
    });
  },
  wrapToolCall: (request, handler) => {
    // Handle execution of the dynamic tool
    if (request.toolCall.name === "calculate_tip") {
      return handler({ ...request, tool: calculateTip });
    }
    return handler(request);
  },
});

const agent = createAgent({
  model: "gpt-5.5",
  tools: [getWeather], // Only static tools registered here
  middleware: [dynamicToolMiddleware],
});

// The agent can now use both getWeather AND calculateTip
const result = await agent.invoke({
  messages: [{ role: "user", content: "Calculate a 20% tip on $85" }],
});

Headless tools

Some tools should run where your user's app runs (typically the browser), not inside the process. Headless tools are tool definitions, which include the name, description, and argument schema, that you register on the server with your agent. The implementation is registered only on the client and executed after a short interrupt/resume handshake.

This is different from ordinary tools whose function body runs on the server, and from server-side tool use where the model provider executes built-in tools remotely.

When to use headless tools

Use them when the work depends on the environment, device, or UI that only exists on the client. For example:

• Browser APIs: Geolocation, IndexedDB, Clipboard, Canvas 2D, file pickers, Battery API, etc.
• Privacy and locality: Data stays on the device (for example, local “memory” in IndexedDB).
• Latency: No extra server round trip for purely local operations.
• Structured, safe effects: Prefer many small, typed tools (for example one tool per canvas primitive) instead of sending arbitrary code to eval.

How the pattern works

In both runtimes, the model sees a normal tool it can call, but the actual execution happens outside the server process.

1. Define the tool with tool({ name, description, schema }) from langchain, metadata and validation only, no server-side runner.
2. Attach the real behavior with .implement(async (args) => { ... }), which returns a headless tool implementation (definition + execute function).
3. Register the definition from step 1 with createAgent or your graph so the model sees the tool in its usual tool-calling loop.
4. Pass the implementation from step 2 to your streaming hook's tools option.

When the model issues a tool call for one of these tools, the run interrupts instead of executing the tool locally. Your app can inspect the payload, perform the action in the right environment (for example a browser, another service, or a human review step), then resume the graph with the tool result. When you use the supported JS SDK hooks, they can detect headless-tool interrupts, run the matching client implementation, and submit the resume command for you.

Use the optional onTool callback to observe lifecycle events (start, success, error) for UI feedback such as spinners or toasts.

Prebuilt tools

LangChain provides a large collection of prebuilt tools and toolkits for common tasks like web search, code interpretation, database access, and more. These ready-to-use tools can be directly integrated into your agents without writing custom code.

See the tools and toolkits integration page for a complete list of available tools organized by category.

Tools from MCP servers

Model Context Protocol (MCP) is an open protocol that standardizes how applications expose tools to language models. Rather than hand-authoring a tool, you connect to an MCP server and adapt the tools it advertises into LangChain tools, ready to pass to an agent like any other tool.

LangChain agents call tools defined on MCP servers through the @langchain/mcp-adapters library, which discovers a server's tools and adapts them into LangChain tools.

For setup, transports, and authentication, see Model Context Protocol (MCP).

Server-side tool use

Some chat models feature built-in tools that are executed server-side by the model provider. These include capabilities like web search and code interpreters that don't require you to define or host the tool logic.

Refer to the individual chat model integration pages and the tool calling documentation for details on enabling and using these built-in tools.