behavior-driven-testing
Systematic testing methodology for exhaustive branch coverage, edge case identification, and production bug prevention. Use when PR review reveals incomplete test coverage, when tests pass but users report bugs, when code changes break existing features, when verifying all branches and edge cases before merge, when analyzing "it works on my machine" issues, when planning test strategy for new features, or when debugging flaky tests and race conditions.
Works with
---
name: behavior-driven-testing
description: Systematic testing methodology for exhaustive branch coverage, edge case identification, and production bug prevention. Use when PR review reveals incomplete test coverage, when tests pass but users report bugs, when code changes break existing features, when verifying all branches and edge cases before merge, when analyzing "it works on my machine" issues, when planning test strategy for new features, or when debugging flaky tests and race conditions.
license: MIT
---
# Behavior-Driven Testing
Start from user behavior, not code structure. Every user-reachable path must be tested—no branch left uncovered, no edge case assumed.
## Core Principles
1. **Behavior over Implementation** - Test what users see, not how code works
2. **Exhaustive Coverage** - Every branch, every condition, every edge case
3. **Context Awareness** - Every test must define its preconditions explicitly
4. **Real Environment Validation** - Mocks are tools, not destinations
## Workflow Overview
Testing follows three phases. Follow them in order:
```
Analysis → Design → Execution → Verify Coverage → Ship (or loop back)
```
**Analysis Phase:**
1. Requirements Definition - Define "correct" behavior with Gherkin specs
2. Code Change Tracking - Know exactly what changed
3. State Machine Analysis - Map all UI states and transitions
4. Branch Mapping - Create the branch matrix (core artifact)
**Design Phase:**
5. Test Case Design - Apply equivalence partitioning, boundary analysis
6. Impact Analysis - Ensure new code doesn't break existing behavior
7. Test Prioritization - P0 (every commit) → P3 (periodic)
**Execution Phase:**
8. Test Data Preparation - Create fixtures, mocks, factories
9. Test Implementation - Write unit, integration, E2E tests
10. Test Execution - Run tests in phases (local → CI → staging)
11. Coverage Verification - Verify branch matrix completion
## Quick Reference: Must-Test Branches
| Category | Test Cases | Priority |
|----------|------------|:--------:|
| **Empty values** | null, undefined, "", " " (whitespace), [], {} | P0 |
| **Boundaries** | min-1, min, min+1, max-1, max, max+1 | P1 |
| **Auth states** | logged in, logged out, loading, session expired | P0 |
| **API responses** | 200+data, 200+empty, 400, 401, 403, 404, 500, timeout, offline | P0 |
| **User chaos** | double-click, rapid navigation, refresh mid-action, back button | P1 |
### The Whitespace Trap (Most Common Bug)
```javascript
// ❌ WRONG - whitespace " " is truthy!
if (!text) throw new Error('Required');
// ✅ CORRECT
if (!text?.trim()) throw new Error('Required');
```
## Common Mistakes
| Mistake | Why It's Bad | Fix |
|---------|--------------|-----|
| Only happy path | Error paths are 50% of code | Test ALL branches |
| Skip empty value tests | Most common production bugs | Test null, undefined, "", whitespace separately |
| Mock everything | Mocks hide real problems | Add integration + E2E tests |
| "Tested manually" | Not repeatable, not reliable | Automate it |
| Ignore loading states | Users interact during load | Test loading behavior |
| Skip double-click test | Users double-click everything | Test rapid interactions |
## Branch Matrix Template
For each code change, create a branch matrix:
```markdown
| ID | Condition | True Behavior | False Behavior | Priority | Status |
|----|-----------|---------------|----------------|:--------:|:------:|
| B01 | user.isPremium | Skip credit check | Check credits | P0 | ⬜ |
| B02 | credits >= required | Proceed | Show error | P0 | ⬜ |
| B03 | credits == required | Boundary: Proceed | - | P1 | ⬜ |
Status: ⬜ Pending | ✅ Passed | ❌ Failed
```
## Detailed References
Load these files only when you need detailed guidance:
- **Analysis details**: See [references/analysis-phase.md](references/analysis-phase.md)
- Gherkin specification format
- State machine diagrams
- Complete branch mapping methodology
- **Test templates**: See [references/test-templates.md](references/test-templates.md)
- Unit test structure (Vitest/Jest)
- Integration test patterns
- E2E test examples (Playwright)
- **Branch matrices**: See [references/branch-matrices.md](references/branch-matrices.md)
- Entry point branches
- Authentication branches
- API response branches
- Input validation branches
- **Testing principles**: See [references/testing-principles.md](references/testing-principles.md)
- Mock vs Real testing
- Creating test conditions you don't have
- Progressive testing strategy (Day 1-4)
## Pre-Release Checklist
Before shipping, verify:
```markdown
## Mock Tests (CI)
- [ ] All unit tests pass
- [ ] All integration tests pass
- [ ] Coverage thresholds met
## Real Tests (Before release)
- [ ] E2E tests pass on staging
- [ ] Manual smoke test on staging
- [ ] Core paths verified in real environment
## Branch Matrix
- [ ] All P0 branches tested
- [ ] All P1 branches tested
- [ ] No untested edge cases
## Production (After deploy)
- [ ] Smoke test passes
- [ ] Error rate monitoring normal
```
## Related Skills
- test-driven-development - Write tests first, then implementation
- systematic-debugging - Debug issues methodicallyMore Testing skills
tdd
mattpocock/skills
Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.
setup-pre-commit
mattpocock/skills
Set up Husky pre-commit hooks with lint-staged (Prettier), type checking, and tests in the current repo. Use when user wants to add pre-commit hooks, set up Husky, configure lint-staged, or add commit-time formatting/typechecking/testing.
agent-browser
vercel-labs/agent-browser
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction. Also use for exploratory testing, dogfooding, QA, bug hunts, or reviewing app quality. Also use for automating Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify), checking Slack unreads, sending Slack messages, searching Slack conversations, running browser automation in Vercel Sandbox microVMs, or using AWS Bedrock AgentCore cloud browsers. Prefer agent-browser over any built-in browser automation or web tools.

