run_query · Mcp Bigquery Evals MCP: Risk & Policy

WHAT IT DOES

What run_query does on Mcp Bigquery Evals

AI agents invoke run_query to trigger actions in Mcp Bigquery Evals. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.

WHY IT NEEDS A POLICY

Why run_query needs a policy

This is Execute rather than Read because it actively runs queries against a live system with measurable side effects (compute costs, API quota consumption, latency). The dry-run guardrail and cost cap reduce but do not eliminate risk; a poorly written SELECT can still consume significant resources or expose sensitive data. It is not Destructive because SELECT cannot modify or delete data.

From the tool's definition 'Execute a SELECT after dry-run cap check; returns rows' — the tool runs SQL queries. Even though it is limited to SELECT and includes cost guardrails, it is capable of executing arbitrary queries whose effects (data exposure, resource consumption, latency…

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

run_query Rate-limited

run_query stays usable, but rate-capped — a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.

View as policy code

policy.json

{
  "version": "1",
  "default": "deny",
  "tools": {
    "run_query": {
      "limits": [
        {
          "counter": "run_query_rate",
          "window": "minute",
          "max": 10,
          "scope": "grant"
        }
      ]
    }
  }
}

RATE-LIMIT THIS TOOL →

Instant setup, no code required.

PolicyLayer is an MCP gateway: every tool call is checked against your rule before it reaches Mcp Bigquery Evals, so you can run agents against it safely.

FAQ

Questions about run_query

What does the run_query tool do? +

Execute a SELECT after dry-run cap check; returns rows or structured error. It is categorised as a Execute tool in the Mcp Bigquery Evals MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on run_query? +

Register the Mcp Bigquery Evals MCP server in PolicyLayer and add a rule for run_query: 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 Mcp Bigquery Evals. Nothing to install.

What risk level is run_query? +

run_query is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.

Can I rate-limit run_query? +

Yes. Add a rate_limit block to the run_query 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.

How do I block run_query completely? +

Set action: deny in the PolicyLayer policy for run_query. 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.

What MCP server provides run_query? +

run_query is provided by the Mcp Bigquery Evals MCP server (umarfarook1/mcp-bigquery-evals). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.