AI agents call transactions_by_category 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.
This tool retrieves financial transaction data categorized by spending category—a read-only query with no side effects. However, severity is elevated to 'high' because it exposes sensitive financial data (transaction history and spending patterns) that could reveal personal financial behavior, income, spending habits, and account details if accessed by unauthorized agents or during prompt injection attacks.
From the tool's definition Tool name 'transactions_by_category' and server description 'query your bank accounts, balances, and transactions through Plaid' indicate data retrieval.
Attacks that exploit this kind of access
transactions_by_category. 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 transactions_by_category: 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.
transactions_by_category 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 transactions_by_category 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 transactions_by_category. 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.
transactions_by_category is provided by the Plaid MCP server (lukew0824/plaid-mcp-server). 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 →