Agent-ready cost guardrail. Returns allow / warn / block with blocking_resources so an AI agent can veto an expensive IaC write before committing. Thresholds cascade: per-call params > CLOUDCOST_GUARDRAIL_* env > CLOUDCOST_BUDGET_* env > no-op allow.
AI agents call check_cost_budget to retrieve information from CloudCost MCP Server without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
check_cost_budget is a guardrail validation function that reads configuration (thresholds from params/env vars) and returns a cost assessment status. It has no side effects—it does not create, modify, delete, or execute infrastructure; it does not transfer funds or charge accounts; it merely advises an agent whether to proceed. This is purely informational/read-class behavior.
From the tool's definition Returns allow/warn/block status with blocking_resources; no modifications, deletions, or financial transactions occur.
Attacks that exploit this kind of access
Agent-ready cost guardrail. Returns allow / warn / block with blocking_resources so an AI agent can veto an expensive IaC write before committing. Thresholds cascade: per-call params > CLOUDCOST_GUARDRAIL_* env > CLOUDCOST_BUDGET_* env > no-op allow. It is categorised as a Read tool in the CloudCost MCP Server MCP Server, which means it retrieves data without modifying state.
Register the CloudCost MCP Server MCP server in PolicyLayer and add a rule for check_cost_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 CloudCost MCP Server. Nothing to install.
check_cost_budget 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 check_cost_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 check_cost_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.
check_cost_budget is provided by the CloudCost MCP Server MCP server (jadenrazo/cloudcostmcp). 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 →