How to Use the Salesforce MCP in LangChain

Q: Can my LangChain agent create new leads in Salesforce?

Yes. As long as the createrecord tool is given to the agent, it can create any SObject, including Leads. The agent just needs to know the object API name ('Lead') and the fields to populate.

Q: Is it possible to run complex SOQL queries with LangChain and Salesforce?

Absolutely. The soqlquery tool accepts any valid SOQL string. Your LangChain agent can construct and execute detailed queries to pull specific Account, Opportunity, or Contact data right into your chain's workflow.

Q: What's the best way to prevent my LangChain agent from deleting Salesforce data?

Don't give it the deleterecord tool. When you initialize the tools for your agent, you can filter the list to exclude any write-heavy or destructive operations. You have full control over the agent's permissions.

Build autonomous Salesforce agents with LangChain by chaining CRM operations into powerful, observable workflows.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers — Included with Plan

Connect Salesforce MCP to LangChain

Create your Vinkius account to connect Salesforce to LangChain — we handle the hosting, security, and runtime updates so you don't have to. No server setup required.

GDPR Included with Plan

Setup Salesforce with LangChain

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Chain CRM Operations Together

This isn't just about calling single tools. It's about building agents that think in steps. Your agent can call `list_reports` to see what's available, then feed a specific report ID into `run_report` to get the data, and finally use that data to perform a targeted `update_record` on an account. Every step is a link in a chain. The output from one tool becomes the input for the next. With LangSmith, you can trace the agent's entire thought process: which tools it chose, what data it passed, and where it spent its time. No more black boxes.

Run Queries and Act on Data

Give your agent the power to explore. It can start with a broad `soql_query` like `SELECT Id, Name FROM Account WHERE IsCustomerPortal = false` to find a set of targets. From there, it can loop through the results, calling `get_record` on each ID to get the full picture. Once it has the data, it can act. This is the core of a ReAct agent. It decides if a record needs a simple `update_record` or if it needs to `create_record` for a new related object, like a follow-up Task. Your agent does the tedious work of fetching and modifying.

Dynamic Record Management for LangChain

Your agent can build its own context. Before attempting to `create_record`, it can call `describe_object` to learn the exact fields and data types required for an Opportunity or a custom object. This cuts down on failed API calls and malformed data. When it needs to find something specific, `global_search` is faster than running multiple queries. If it finds a duplicate record, it can flag it for review or, if you've given it the permission, use `delete_record` to clean it up. This MCP server gives your agent the tools to manage data, not just read it.

Setup guide

Set up Salesforce MCP in LangChain

Prerequisites

Python 3.10+ installed
langchain-mcp-adapters + langgraph packages
Active Vinkius subscription with a valid endpoint token

1

Install dependencies
Run pip install langchain-mcp-adapters langgraph langchain-openai. The MCP adapters package converts MCP tools into native LangChain BaseTool objects.
2

Connect via HTTP transport
Use MultiServerMCPClient with "transport": "http" pointing to your Vinkius endpoint. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com.
3

Create a ReAct agent
Pass the discovered tools to create_react_agent() from LangGraph. The agent automatically routes Salesforce tool calls through the MCP protocol.
4

Run with any LLM
Swap ChatOpenAI for ChatAnthropic, ChatGoogleGenerativeAI, or any LangChain-compatible model. The MCP tools work identically across all providers.

agent.py

from langchain_mcp_adapters.client import MultiServerMCPClient
from langgraph.prebuilt import create_react_agent
from langchain_openai import ChatOpenAI

async with MultiServerMCPClient({
    "salesforce-mcp": {
        "transport": "http",
        "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp",
    }
}) as client:
    tools = client.get_tools()

    agent = create_react_agent(
        ChatOpenAI(model="gpt-4o"),
        tools,
    )
    result = await agent.ainvoke({
        "messages": "List recent Salesforce transactions"
    })
    print(result["messages"][-1].content)

Get your connection token →

Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Salesforce. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.

Why Choose Vinkius

Vinkius connects your tools to AI with real-time monitoring and automatic cost savings — all from one dashboard.

Connect Salesforce now

Real-time monitoring

Live

visibility into every interaction

Connect your favorite tools to your AI and see exactly what's happening — every request, every response, in real time.

Built-in savings

60%

lower AI costs

Vinkius compresses data between your apps and your AI automatically. Lower bills every month — no configuration required.

Single dashboard

One

place for every integration

Every tool your AI connects to, managed from a single screen. One account, complete control.

Common questions about Salesforce MCP in LangChain

You'll use the LangChain MCP adapter. Just instantiate the client with your server URL, call `client.get_tools()`, and pass the resulting tool list directly into your agent executor. It's designed to work out of the box.

Yes. As long as the `create_record` tool is given to the agent, it can create any SObject, including Leads. The agent just needs to know the object API name ('Lead') and the fields to populate.

Absolutely. The `soql_query` tool accepts any valid SOQL string. Your LangChain agent can construct and execute detailed queries to pull specific Account, Opportunity, or Contact data right into your chain's workflow.

Don't give it the `delete_record` tool. When you initialize the tools for your agent, you can filter the list to exclude any write-heavy or destructive operations. You have full control over the agent's permissions.

This server acts as a direct, stateless pass-through. When your agent requests data like Contacts or Accounts, it flows from Salesforce through our ephemeral sandbox to your LangChain application and is gone. We don't log or store your CRM data. Your Vinkius token secures the connection; we never see your Salesforce credentials.

Use it with your favorite AI tools

Connect this server to Cursor, Claude, VS Code, and more.

OpenAI Agents SDK sdk-python

Google ADK sdk-python

Pydantic AI sdk-python

Vercel AI SDK sdk-typescript