AI agents use update_category_group to create or update resources in Ynab — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Ynab environment.
This tool modifies budget metadata (category group structure) reversibly. It does not delete data (ruling out Destructive), does not move money (ruling out Financial), and does not execute arbitrary code (ruling out Execute). The sibling context shows both Write operations (create_*) and Destructive operations (delete_*), confirming this falls in the Write category.
From the tool's definition Tool name 'update_category_group' indicates modification of existing data. Description states 'Update a category group'. In the YNAB financial management context, category groups organize budget categories.
Attacks that exploit this kind of access
[1 API call] Update a category group. It is categorised as a Write tool in the Ynab MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Ynab MCP server in PolicyLayer and add a rule for update_category_group: 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 Ynab. Nothing to install.
update_category_group 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 update_category_group 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 update_category_group. 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.
update_category_group is provided by the Ynab MCP server (justmytwospence/ynab-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 →