Artificial intelligence news.
Straight from the original announcements.

LangChain launches Tool Gateway for governed agent tool use

November 19, 2026, 16:00By the AI News Bites Editorial TeamSource: LangChain
LangChain launches Tool Gateway for governed agent tool use
Image: LangChain

LangChain has launched LangSmith Tool Gateway in Public Beta, giving teams a managed layer for controlling how agents access tools and credentials. The feature is available on the LangSmith Agent Platform and is designed to support access management, approvals and auditing for agent tool calls.

Key points
  • LangChain launched LangSmith Tool Gateway in Public Beta.
  • The service manages tool access, identity and policies through a single URL.
  • Tool kits restrict agents to approved tools, with search mode for large collections.
  • Administrators can configure deny and human approval policies.
  • Tool calls appear in LangSmith traces with identity, server, outcome and duration details.

A managed layer for agent tools

Tool Gateway routes tool requests through a single URL. For each caller, it resolves identity, manages access and logs tool call traces, according to LangChain.

Teams can register MCP servers by URL or enable built-in servers for common apps such as Google, Salesforce and GitHub. They can then create curated tool kits, which bundle selected tools at a single URL. When an agent connects to a tool kit, it sees only the tools included in that kit.

LangChain also provides a search mode for large tool collections. Instead of loading hundreds or thousands of tools at once, an agent can query its kit and load tools relevant to the current task.

Identity and policies for every call

Tool Gateway supports two identities for tool calls. An agent can act as the user, using credentials connected through an identity provider such as Okta or Entra ID. If the user has not connected an account, the call returns a sign-in link and continues after authentication.

An agent can also act with credentials tied to the agent itself. LangChain says these credentials can be configured through agent authentication in LangSmith or through profiles. Profiles use shared credentials that a workspace administrator connects once for agents that run autonomously in the background.

Administrators can set deny policies at the organization or workspace level. These can apply to specific users, service keys or combinations of them, as well as to everyone with listed exceptions. Approval policies can require a user to confirm a sensitive call in their MCP client before it runs, such as a call that writes to a data warehouse.

Tracing tool use in LangSmith

Every tool call appears in a LangSmith agent trace. With Tool Gateway, the trace includes information such as the credentials used, the server, the user, the outcome and the duration. LangChain says the tool's input and output data are not stored in the record.

The company describes the feature as part of a broader governance approach alongside LangSmith's LLM Gateway. LLM Gateway governs model calls with capabilities including spend limits, model fallbacks, rate limits and redaction of personal data and secrets. Together, the two gateways are intended to provide visibility into agent actions and spend.

LangChain illustrates the need with a large US retailer building agents for store operations and inventory management. The retailer has built more than 50 MCP servers with more than a thousand tools, while individual agents need only approved subsets and must use the right credentials for each context. Tool Gateway is intended to centralize those controls rather than require teams to build authentication and routing separately for every agent.

Related news

LangSmith adds alerts for errors, latency and LLM quality issues
For developersNovember 18, 2026, 20:33

LangSmith adds alerts for errors, latency and LLM quality issues

LangChain has launched Alerts in LangSmith, giving developers a way to monitor LLM applications and agents for production problems in real time. The feature can flag…

LangChain adds cross-agent tracking to help control coding costs
For developersNovember 18, 2026, 20:33

LangChain adds cross-agent tracking to help control coding costs

LangChain is expanding LangSmith to give teams a shared view of spending and activity across multiple coding agents. The company says the tools can help organizations…

LangSmith links coding agent spending to pull request outcomes
Products and toolsNovember 18, 2026, 16:00

LangSmith links coding agent spending to pull request outcomes

LangChain has added coding agent reporting to LangSmith, linking spending to GitHub pull requests and GitLab merge requests. The company says the feature is designed to…