Run a large SELECT and stream rows to a NDJSON file on disk — for exports that exceed execute_query
AI agents call streaming_query to retrieve information from Querybridge without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The core operation is a SELECT query, which is a read operation with no side effects on the database itself. Although the tool writes results to a file on disk (a potential write concern), the primary action and security risk derives from data retrieval and export capability.
From the tool's definition Tool explicitly performs 'large SELECT' and 'stream rows', both read-only operations on a database. No modification, deletion, or execution capability is described. However, the output is written to disk, which persists data outside the database.
Attacks that exploit this kind of access
Run a large SELECT and stream rows to a NDJSON file on disk — for exports that exceed execute_query. It is categorised as a Read tool in the Querybridge MCP Server, which means it retrieves data without modifying state.
Register the Querybridge MCP server in PolicyLayer and add a rule for streaming_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 Querybridge. Nothing to install.
streaming_query 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 streaming_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.
Set action: deny in the PolicyLayer policy for streaming_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.
streaming_query is provided by the Querybridge MCP server (mahmoudhassanmustafa/querybridge-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →