AI agents call plaid_query to retrieve information from Plaid without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
While this is a Read operation (no data modification or deletion), the severity is elevated to 'high' because it provides direct access to sensitive financial data including account balances, transaction history, and holdings. Unauthorized or careless access via an AI agent could expose private financial information, enable fraud detection evasion, or support social engineering.
From the tool's definition Tool description explicitly states 'read-only SQL query' with access to 'financial accounts', 'transactions, balances, and holdings'. The name 'plaid_query' and description both confirm retrieval-only semantics without modification capabilities.
Attacks that exploit this kind of access
Run a read-only SQL query against your Plaid-linked financial data. It is categorised as a Read tool in the Plaid MCP Server, which means it retrieves data without modifying state.
Register the Plaid MCP server in PolicyLayer and add a rule for plaid_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 Plaid. Nothing to install.
plaid_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 plaid_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 plaid_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.
plaid_query is provided by the Plaid MCP server (yuechen/plaid-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 →