Confidence Scoring

TLDR

After every task, rate your confidence from 1-10. If it's below 8, fix it before moving on.

This one habit prevents the cascade failure where broken foundations corrupt everything built on top.

---

How It Works

When you finish a task, ask yourself (or have AI ask itself):

"How confident am I that this is done right?"

8/10 means: It works, it's tested, errors are handled, code is clean. Solid enough to build on.

9/10 means: All of the above, and it's genuinely good work. You'd show it to someone.

7/10 means: Something's off. Maybe tests are thin, edge cases unhandled, or you're not sure about a decision. Fix it.

6/10 or below: Stop. Something is fundamentally wrong.

---

The Format

Keep it simple:

## Confidence: 8/10

**Must-haves (met):**
- [x] Core functionality works
- [x] Errors handled
- [x] Tests pass
- [x] Code readable

**Deferred:**
- [ ] Rate limiting (Phase 2)
- [ ] Advanced validation (v1.0)

That's it. 10 lines. Don't over-engineer this.

---

What Counts as 8/10

The bar isn't perfection. It's "solid enough to build on."

8/10 code:

• Does what it's supposed to do
• Handles the obvious error cases
• Has tests for the important paths
• Is readable by someone else (or AI next session)
• Doesn't have known bugs you're ignoring

8/10 is NOT:

• Every edge case handled
• 100% test coverage
• Perfectly optimized
• Production-hardened

Those are 9/10 or 10/10. For MVP, 8/10 is the target.

---

When You're Below 8

Don't negotiate with yourself. If you're at 7/10, something's wrong.

Common gaps:

• "Tests pass but I'm not sure they test the right things" → Write better tests
• "It works but error handling is sketchy" → Add error handling
• "I'm not sure about this architectural decision" → Think it through or ask
• "There's a bug I haven't fixed" → Fix the bug

The conversation:

AI: "Confidence: 7/10. The login works but I haven't 
tested what happens when the database is unavailable."

You: "Add that test and handling before we continue."

AI: [adds test, adds handling]

AI: "Confidence: 8/10. Database errors now return 503 
with retry guidance. Test added."

Now you can move on.

---

Why This Matters

Without confidence scoring, broken code accumulates:

Task 1: Auth (buggy) → "good enough, continue"
Task 2: Database (built on buggy auth) → "some issues, continue"  
Task 3: API (built on buggy database) → "not sure why this fails"
Task 4: UI (nothing works) → "everything is broken"

With confidence scoring:

Task 1: Auth → 7/10 → Fix → 8/10 → Continue
Task 2: Database → 8/10 → Continue
Task 3: API → 8/10 → Continue
Task 4: UI → Works because foundation is solid

The extra hour fixing Task 1 saves days debugging Task 4.

---

Honest Self-Assessment

The scoring only works if you're honest.

Temptation: "It's probably fine, I'll call it 8/10 and move on."

Reality check: Would you bet $100 that this task is solid? If not, it's not 8/10.

AI tends to be optimistic. When AI says 8/10, verify. Run the tests. Try the edge cases. Check the error handling yourself.

---

Quick Reference

Score	Meaning	Action
9-10	Excellent	Continue, maybe celebrate
8	Solid	Continue
7	Gaps exist	Fix before continuing
6	Significant issues	Stop and fix
≤5	Fundamentally broken	Reconsider approach

---

Next: Phase Audits — Fresh eyes between major milestones.