AI agents use update_budget to create or update resources in Flour Ads — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Flour Ads environment.
This tool modifies an existing campaign's budget parameter, which is a reversible write operation. While it affects spending, it is classified as Write rather than Financial because it adjusts a budget limit rather than directly executing payment transactions. However, severity is high because misuse could cause significant unintended ad spend. The confidence is high given the clear naming and description provided.
From the tool's definition Tool name 'update_budget' and description 'Update the daily budget for a campaign' indicate modification of campaign budget settings.
Attacks that exploit this kind of access
Update the daily budget for a campaign. It is categorised as a Write tool in the Flour Ads MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Flour Ads MCP server in PolicyLayer and add a rule for update_budget: 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 Flour Ads. Nothing to install.
update_budget 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_budget 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_budget. 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_budget is provided by the Flour Ads MCP server (wayy-research/flour-ads). 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 →