Testing Guide - HLMS Docs

Test Suites Overview

HLMS has 112 tests across 6 suites:

Suite	Name	Tests	Focus
A	Entry Points & Navigation	7	Login, routing, basic navigation
B	Student Workflow	35	Dashboard, assignments, submission
C	Teacher Workflow	40	Review, grading, feedback
D	Notification System	16	Bidirectional notifications
E	Data Persistence	10	localStorage, data integrity
F	Edge Cases & Error Handling	4	Error states, edge cases

Testing Modes

1. Smoke Mode (7 tests, 10-15 min)

Quick sanity check before commits.

/testing-agent smoke

Tests: A1, B2, C2, D2, E2, F1, F2

2. Diff Mode (dynamic)

Only test changed files based on git diff.

/testing-agent diff

3. Risky Mode (filtered)

Test high-risk areas only based on risk scores.

/testing-agent risky

4. Full Mode (112 tests, 45-60 min)

Complete test suite - all 6 suites.

/testing-agent full

5. Suite Mode (7-40 tests)

Run specific suite only.

/testing-agent suite B

Project Suites

Project suites group test suites for major releases:

`HLMS-1.0-FULL`	All 112 tests (Suites A-F)
`HLMS-1.0-CORE`	~35 critical path tests
`HLMS-1.0-UI`	User interface tests (planned)
`HLMS-1.0-DATA`	Data layer tests (planned)

/testing-agent run HLMS-1.0-CORE
/testing-agent coverage HLMS-1.0-FULL

Key Test Cases

Critical Path Tests

A1 - Login page loads
B2 - Student dashboard renders
B6 - Assignment submission works
C2 - Teacher portal loads
C6 - Grading saves correctly
D2 - Notifications appear
E2 - Data persists after refresh

Test Dependencies

A1 (Navigation)
  → B2 (Student Dashboard)
     → B6 (Assignment Submission)
        → C2 (Teacher Dashboard)
           → C6 (Teacher Grading)
              → D2 (Notifications)
                 → E2 (Data Persistence)

Confidence Scoring

4-factor scoring system (40-100%):

Static Analysis	40 points	Code structure, syntax
Historical Pattern	30 points	Known bug patterns
Test Correlation	20 points	Direct test coverage
Code Context	10 points	Surrounding code quality

Confidence Levels

95-100% - Certain (auto-fix safe)
80-94% - High (recommend fix)
60-79% - Medium (review needed)
40-59% - Low (manual investigation)
0-39% - Speculative (don't use)

Risk Scoring

Predictive risk for code changes (1-10 scale):

Risk = (Coupling × 0.35) + (Churn × 0.25) + (Complexity × 0.20) + (History × 0.20)

Current High-Risk Files

🔴 `student-assignments-enhanced.html`	8.05
🟠 `teacher-review.html`	7.00
🟠 `data-manager.js`	6.55
🟡 `student-dashboard.html`	4.20
🟢 `submission-review.html`	2.30

Deployment Gates

Risk < 3.0 - 🟢 Deploy anytime
Risk 3.0-5.9 - 🟡 Deploy with smoke test
Risk 6.0-7.9 - 🟠 Deploy with full test suite
Risk ≥ 8.0 - 🔴 Requires manual review + full testing

Running Tests

# Smoke test before commit
/testing-agent smoke

# Full test after major changes
/testing-agent comprehensive

# Specific suite
/testing-agent suite C

# Check coverage
/testing-agent coverage HLMS-1.0-FULL

# View risk scores
/testing-agent risk-report

Testing Agent Files

Located in .claude/agents/:

testing-agent-v2.md - Main agent config
confidence-scoring.md - Confidence system
regression-risk-scoring.md - Risk analysis
project-suites.md - Suite definitions
pattern-changelog.md - Known patterns
test-dependencies.md - Test cascade map
coverage-gap-detector.md - Coverage analysis