Implementation brief

## Who Uses It
A deal desk preparing custom product quotes.

## What You Will Build
Build an agent that converts a sales request into a calculated quote with product lines, discounts, dates, and approval requirements. Let reviewers adjust the quote and inspect every pricing decision.

## Inputs
- Sales request
- Product price list
- Proration and approval policies

## Workflow
1. Extract the requested products, quantities, and contract dates.
2. Calculate line totals using code and supplied pricing rules.
3. Identify discount or margin exceptions.
4. Generate a versioned quote and approval request.

## Interface
A quote editor with line calculations, policy explanations, approval status, and an exportable quote.

## Working Demo
Produce a quote with list price 3600, discounted price 2880, and a pending approval for the 20 percent discount.

## Deliverables
- A working local application with the specified review interface.
- Source code, dependency lockfile, and setup commands.
- A recorded end-to-end demo using the supplied starter data.
- Example outputs, action logs, and automated tests.

## Starter Data
These fictional records are a small development fixture, not a production dataset. Implement the mock services described in the brief.

```json
{
  "request": {
    "product": "Support Suite",
    "seats": 10,
    "months": 12,
    "requested_discount_percent": 20
  },
  "price_list": [
    {
      "product": "Support Suite",
      "monthly_seat_price": 30
    }
  ],
  "policy": {
    "maximum_unapproved_discount_percent": 10
  },
  "approval": null
}
```

## Constraints
Use fictional pricing rules. Do not issue binding quotes or infer real tax treatment.
Choose an available model that supports your implementation. Record its identifier and version.

## Evaluation
Review the complete application, not only the starter-case answers.
- The quote shows list and discounted totals.
- The discount exception requires approval.
- Changing the quote invalidates approval of an older version.
Add one normal case, one incomplete-input case, and one failure-recovery case.