Skip to content
← Projects

Mandate

Spending guardrails for AI agents: the LLM proposes, code decides

Domain
AI Safety
Status
Prototype
Year
2026
Stars
★ 3
Built for
PayPal AI Hackathon 2026

Mandate lets a team set spending rules that an AI shopping agent must operate within. The LLM only proposes products and quantities; a deterministic TypeScript policy engine decides ALLOW, NEEDS_APPROVAL or DENY, and every decision is written to an append-only audit log. Built for the PayPal AI Hackathon 2026 (sandbox only).

The problem

Giving an AI agent a payment method is currently all-or-nothing, exposing teams to overspend, price drift, prompt injection, structuring and lack of accountability.

How it works

  1. 01

    Treasurer defines a mandate: budget, per-item cap, categories, approval threshold

  2. 02

    LLM parses plain-language requests and ranks catalog items, never handling prices

  3. 03

    Deterministic policy engine recomputes prices and returns ALLOW / NEEDS_APPROVAL / DENY

  4. 04

    Payments via PayPal Orders v2 sandbox; human approvals via Zapier

  5. 05

    Dashboard (AG Grid) over an append-only audit log

Watch

Mandate — demo

Same domain