AI agents call read_dataset to retrieve information from LangSmith MCP Server without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves dataset information from LangSmith without modifying, deleting, or executing operations. The 'read_' prefix and context of observability data access indicate it queries and returns existing data. Confidence is slightly reduced (0.85 rather than higher) due to empty tool description, but the naming convention and server purpose make the classification clear.
From the tool's definition Tool name 'read_dataset' indicates a retrieval operation. The server description mentions 'working with datasets' as part of observability platform access.
Documented attack patterns abuse exactly the kind of access read_dataset gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and LangSmith MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for read_dataset:
{
"version": "1",
"default": "deny",
"tools": {
"read_dataset": {}
}
}read_dataset is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
read_dataset. It is categorised as a Read tool in the LangSmith MCP Server MCP Server, which means it retrieves data without modifying state.
Register the LangSmith MCP Server MCP server in PolicyLayer and add a rule for read_dataset: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches LangSmith MCP Server. Nothing to install.
read_dataset is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the read_dataset rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for read_dataset. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
read_dataset is provided by the LangSmith MCP Server MCP server (langchain-ai/langsmith-mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from LangSmith MCP Server, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
15 LangSmith MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.