Unlock agent spending without losing control.

Let AI agents pay for the data, tools, compute, and services they need. Lootrunners keeps spending inside the budgets, rules, vendors, and approvals you set, then shows you exactly what every dollar bought.

Start read-only. Grant spending authority only when you're ready.

Research Agent spent $19 on 2,000 records for Midwest Expansion. Maya approved the new vendor once. 1,842 records were accepted. The payment was reconciled.

Know what every dollar paid for.

Scout investigates spending across the agents, wallets, chains, projects, policies, and accounting records your team can access. Ask questions in plain English and get the transactions, calculations, and evidence behind the answer.

Scout calculates through governed query tools and links every financial claim back to evidence. It cannot increase an agent's authority.

Turn a question into a number you can trust.

Scout can generate charts and comparisons from your financial history while keeping the calculation behind them visible: the wallets, agents, chains, filters, time range, metric definition, data freshness, and query lineage behind every answer.

See exactly what the agent knew when it decided to spend. Playback reconstructs the sequence without rewriting history when new information appears later.

Agents can request spending. They cannot give themselves more authority. Critical limits can be enforced below the application layer through supported wallet and provider controls.

Compliance obligations depend on the customer's activities and jurisdictions. Lootrunners provides configurable controls and evidence, not a blanket legal-compliance guarantee.

Know what the work actually cost.

Put APIs, compute, data, assets, contractors, fees, and other costs against the same project or outcome. Know the cost of the result, not just the cost of the model.

Agents request purchases through REST, MCP, or SDK tools and provide the business context behind the request. Lootrunners checks it against the agent's Mandate and returns a decision — automatic, needs approval, or blocked — then keeps the evidence and outcome connected.

const res = await fetch('https://app.lootrunners.com/api/agent/v1/spend/intents', {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${LOOTRUNNERS_AGENT_KEY}`,
    'Idempotency-Key': `lead-${lead.id}`,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    mandate_id: 'midwest-research',
    purpose: 'Verify employer and title',
    amount_usd: '0.08',
    vendor_domain: 'example-data.com',
  }),
});

// decision.result is AUTO, ASK, or BLOCK
const { decision } = await res.json();
return decision.result;