Spend Control for Trae
Trae ships AI-native editing with agentic coding. Agents that code can also buy — the guard makes the budget real.
Why Trae agents overspend
Agentic sessions consume usage and call tools fast.
Paid tool calls happen mid-task without review.
How it works
Register the guard as a tool: before any spend, the agent gets APPROVED, BLOCKED, or FLAGGED.
Rules that fit
Per-transaction cap, merchant allowlist, velocity limit.
Guard tool
const res = await fetch("https://sipi.bot/v1/transactions/evaluate", {
method: "POST", headers: { "Content-Type": "application/json" },
body: JSON.stringify({ amount: 60, merchant: "gpu-provider.com", category: "compute" }),
});
const { decision } = await res.json();One tool, enforced policy.
FAQ
Does this slow Trae?
No — ~5 ms per check, spend actions only.
Is this like the Cursor integration?
Same pattern — see the Cursor page.
Related
Stop the next $12,400 night.
One API call (or MCP tool) in front of every agent transaction — APPROVED, BLOCKED, or FLAGGED, deterministic, ~5 ms, fully logged.
See plans — from $99/mo Try a live check