Back to Prompt Library
Build the complete application
Implementation brief
Inspect the original prompt language first, then copy or adapt it once you know how it fits your workflow.
Linked challenge: Build a Customer Support Resolution Agent
Format
Code-aware
Lines
52
Sections
11
Linked challenge
Build a Customer Support Resolution Agent
Prompt source
Original prompt text with formatting preserved for inspection.
52 lines
11 sections
No variables
1 code block
## Who Uses It
A support team managing requests that span orders, billing, and account access.
## What You Will Build
Build an agent that investigates a customer ticket across order and account records, proposes a resolution, and routes exceptions for approval. Give support staff one workspace to review actions and customer replies.
## Inputs
- Customer ticket
- Order and account records
- Support policies
## Workflow
1. Verify the customer and retrieve relevant records.
2. Separate the request into actionable tasks.
3. Resolve permitted tasks in mock systems and escalate exceptions.
4. Draft a customer response with an accurate case status.
## Interface
A ticket workspace showing evidence, completed tasks, pending approvals, and an editable response draft.
## Working Demo
Retrieve tracking, draft the delay explanation, escalate the refund, and keep the case open until approval.
## 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
{
"ticket": {
"id": "T1",
"customer_id": "C1",
"text": "My order is late. Please refund 80 and tell me where it is."
},
"order": {
"id": "O1",
"customer_id": "C1",
"status": "delayed",
"tracking": "TR1"
},
"policy": {
"agent_refund_limit": 50
},
"approval": null
}
```
## Constraints
Use mock orders and billing. Do not send messages or issue real refunds.
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 response cites order O1 and tracking TR1.
- The refund request is routed to a human.
- The unresolved case is not marked closed.
Add one normal case, one incomplete-input case, and one failure-recovery case.Adaptation plan
Keep the source stable, then change the prompt in a predictable order so the next run is easier to evaluate.
Keep stable
Preserve the source structure until you know which part of the prompt is actually driving the result quality.
Tune next
Change domain facts, examples, and tool context first before you rewrite the instruction scaffold.
Verify after
Validate one failure mode at a time so prompt changes stay attributable instead of getting noisy.