Scenario 01: New Ticket Triage and Classification¶

Scenario Overview¶

Pre-conditions¶

• Workspace open with all folders visible
• No existing work-item folder for NTK-10005
• Mock JIRA client script available at scripts/mock-jira-client.py
• Virtual environment activated

Exact Prompt to Use¶

"I need to work on a new ticket. Query our JIRA board for new tickets, find NTK-10005 (Wristband RFID Field), and create the initial workspace with ticket report and simple explanation. Classify whether this is architecturally relevant or a code-level task."

Expected AI Actions¶

1. Run python scripts/mock-jira-client.py --list --status "New" to list available tickets
2. Run python scripts/mock-jira-client.py --ticket NTK-10005 to extract full ticket details
3. Create folder structure:

   work-items/tickets/_NTK-10005-wristband-rfid-field/
   ├── NTK-10005-solution-design.md
   ├── 1.requirements/
   │   └── NTK-10005.ticket.report.md
   ├── 2.analysis/
   │   └── simple.explanation.md
   └── 3.solution/
       ├── a.assumptions/assumptions.md
       ├── c.current.state/investigations.md
       ├── d.decisions/decisions.md
       ├── g.guidance/guidance.md
       ├── i.impacts/impacts.md
       └── s.user.stories/user-stories.md
   

4. Generate ticket report from JIRA mock data (not fabricated)
5. Write simple explanation suitable for non-technical stakeholders
6. Classify the ticket as architecturally relevant or code-level task with reasoning

What to Watch For¶

• Does the AI run the mock scripts or fabricate data?
• Does the folder name start with underscore and use kebab-case?
• Does the ticket report faithfully reflect mock data?
• Is classification reasoning grounded in the ticket description?
• Does the simple explanation avoid technical jargon?

Quality Rubric¶

Score each criterion 1-5:

• [ ] Tool Usage: Correctly used mock tool scripts (not guessed data)
• [ ] Folder Structure: Created proper folder structure (underscore prefix, kebab-case)
• [ ] Data Accuracy: Ticket report matches mock data accurately
• [ ] Classification: Classification reasoning is sound and justified
• [ ] Clarity: Simple explanation is clear and non-technical friendly

Maximum Score: 25

Token Cost Tracking¶

Notes¶

• This is the simplest scenario and serves as a baseline
• If an AI tool cannot handle this scenario well, more complex scenarios are unlikely to succeed
• Pay attention to whether the AI follows the naming conventions without being reminded