SOLO CLI Project: Build Something You Actually Want

Goal: Use every SOLO loop to build a small Python CLI project of your choice from scratch.

Pick a project

Choose something you'd actually use. Here are ideas, but feel free to design your own — just keep it CLI-based (no web stuff or databases yet; save that for the next courses):

1. Grade Calculator — enter assignments, weights, and scores per class; see current grade, and predict "what score do I need on the final to get an A?"
2. Flashcard Quizzer — import questions from a text file, quiz yourself, track which ones you get wrong most often, and prioritize those.
3. Budget Tracker — log expenses with categories, see spending summaries by week/month/category, and set alerts for budget limits.
4. Workout Logger — log exercises with sets/reps/weight, see progress over time, and track personal records per exercise.

---

The Full SOLO Sequence

Phase 1: Plan

• Write your specification file (SPEC.md) before writing any code.
• Step A: Plan. Run /architect to get a technical plan (use #file:workflows/architect.md).
• Step B: Critique. Run /critique — label every finding as 🔴/🟡/⚪ (use #file:workflows/critique.md).
• Step C: Revise. Run /revise — address all 🔴 and chosen 🟡 findings (use #file:workflows/revise.md).
• Run /critique again for a second pass.
• Define your interfaces (interfaces.md) showing your function signatures.
• Write your automated test definitions (before implementation).
• Commit everything.

Phase 2: Build

• Step D: Implement. Run /code to write the code (tests should go green) (use #file:workflows/code.md).
• Add explicit input validation for every user input.
• Run tests: pytest -v — everything green, no warnings.
• Run /memory at the end of a session to save context (use #file:workflows/memory.md).
• Start a new session, run /resume, and keep building (use #file:workflows/resume.md).
• If you hit a non-obvious bug, use /sherlock before patching (use #file:workflows/sherlock.md).
• Final pytest -v check — 100% passing.
• Commit everything, push to GitHub.

---

Project Deliverables

All of these must be pushed to your GitHub repository:

1. Working program — a clean script someone can clone and run.
2. SPEC.md — your handwritten specification.
3. PLAN_v2.md — the revised technical architecture plan.
4. CRITIQUE_v1.md — with your critique findings labeled by severity.
5. interfaces.md — your function signatures.
6. Tests — with at least 15 automated test functions.
7. At least one SESSION_MEMORY.md.
8. Post-Mortem (POSTMORTEM.md):
   • Where did SOLO save you time vs. just asking the AI to build it?
   • Which pillar had the biggest impact on code quality?
   • What would you do differently next time?