Add a Costco product to the cart
AI agents use add_to_cart to create or update resources in Mcp Costco — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Mcp Costco environment.
Adding items to a cart modifies user data (the cart contents) but is not destructive (removable), not financial (no payment yet), and not execute (no arbitrary code/commands). It ranks as Write. Severity is medium because misuse could lead to unwanted cart modifications or fraudulent purchases, but the damage is limited to a single user's cart and the transaction is not finalized until checkout.
From the tool's definition Tool description states 'Add a Costco product to the cart', which creates or modifies cart state. This is a write operation that adds items to a shopping cart—reversible via removal.
Attacks that exploit this kind of access
Add a Costco product to the cart. It is categorised as a Write tool in the Mcp Costco MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Mcp Costco MCP server in PolicyLayer and add a rule for add_to_cart: 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 Costco. Nothing to install.
add_to_cart is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the add_to_cart 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 add_to_cart. 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.
add_to_cart is provided by the Mcp Costco MCP server (markswendsen-code/mcp-costco). 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 →