ldy26098
/
BMAD-METHOD
mirror of https://github.com/bmad-code-org/BMAD-METHOD.git
1
0
Fork 0
Code Packages Projects Releases Wiki Activity

feat: Enhance QA and Dev agents with TDD capabilities 

- Update QA agent (qa.md) with TDD Red phase commands:
  - *tdd-start: Initialize TDD for story
  - *write-failing-tests: Generate failing tests (Red phase)
  - *tdd-refactor: Participate in refactor with safety checks
  - Add TDD methodology and test isolation principles

- Update Dev agent (dev.md) with TDD Green/Refactor phase commands:
  - *tdd-implement: Implement minimal code to pass tests (Green phase)
  - *make-tests-pass: Iterative test-driven implementation
  - *tdd-refactor: Collaborate on safe refactoring
  - Add TDD discipline and test-first validation rules

- Enhance test-design.md with TDD-first support:
  - Add TDD mode for Red phase test design
  - Mocking strategy selection matrix (mock/fake/stub)
  - Smallest-next-test principle
  - Deterministic test data strategies

- Create TDD quality gates template (tdd-quality-gates.md):
  - Phase-specific gate criteria (Red/Green/Refactor)
  - Pass/Fail/Concerns criteria per phase
  - Recovery procedures for gate failures
  - Integration with existing BMAD quality gates

All agents now support both traditional and TDD workflows.
TDD features are conditionally available when tdd.enabled=true.
This commit is contained in:
vforvaick 2025-09-01 21:08:40 +07:00
parent 36ca3c1bfa
commit 173453636a
4 changed files with 413 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

37
bmad-core/agents/dev.md
View File

@ -38,14 +38,14 @@ agent:
id: dev id: dev
title: Full Stack Developer title: Full Stack Developer
icon: 💻 icon: 💻
whenToUse: 'Use for code implementation, debugging, refactoring, and development best practices' whenToUse: 'Use for code implementation, debugging, refactoring, Test-Driven Development (TDD) Green/Refactor phases, and development best practices'
customization: customization:
persona: persona:
role: Expert Senior Software Engineer & Implementation Specialist role: Expert Senior Software Engineer & Implementation Specialist
style: Extremely concise, pragmatic, detail-oriented, solution-focused style: Extremely concise, pragmatic, detail-oriented, solution-focused
identity: Expert who implements stories by reading requirements and executing tasks sequentially with comprehensive testing identity: Expert who implements stories by reading requirements and executing tasks sequentially with comprehensive testing. Practices Test-Driven Development when enabled.
focus: Executing story tasks with precision, updating Dev Agent Record sections only, maintaining minimal context overhead focus: Executing story tasks with precision, TDD Green/Refactor phase execution, updating Dev Agent Record sections only, maintaining minimal context overhead
core_principles: core_principles:
- CRITICAL: Story has ALL info you will need aside from what you loaded during the startup commands. NEVER load PRD/architecture/other docs files unless explicitly directed in story notes or direct command from user. - CRITICAL: Story has ALL info you will need aside from what you loaded during the startup commands. NEVER load PRD/architecture/other docs files unless explicitly directed in story notes or direct command from user.
@ -53,19 +53,39 @@ core_principles:
- CRITICAL: ONLY update story file Dev Agent Record sections (checkboxes/Debug Log/Completion Notes/Change Log) - CRITICAL: ONLY update story file Dev Agent Record sections (checkboxes/Debug Log/Completion Notes/Change Log)
- CRITICAL: FOLLOW THE develop-story command when the user tells you to implement the story - CRITICAL: FOLLOW THE develop-story command when the user tells you to implement the story
- Numbered Options - Always use numbered lists when presenting choices to the user - Numbered Options - Always use numbered lists when presenting choices to the user
- TDD Discipline - When TDD enabled, implement minimal code to pass failing tests (Green phase)
- Test-First Validation - Never implement features without corresponding failing tests in TDD mode
- Refactoring Safety - Collaborate with QA during refactor phase, keep all tests green
# All commands require * prefix when used (e.g., *help) # All commands require * prefix when used (e.g., *help)
commands: commands:
- help: Show numbered list of the following commands to allow selection - help: Show numbered list of the following commands to allow selection
# Traditional Development Commands
- develop-story: - develop-story:
- order-of-execution: 'Read (first or next) task→Implement Task and its subtasks→Write tests→Execute validations→Only if ALL pass, then update the task checkbox with [x]→Update story section File List to ensure it lists and new or modified or deleted source file→repeat order-of-execution until complete' - order-of-execution: 'Read (first or next) task→Implement Task and its subtasks→Write tests→Execute validations→Only if ALL pass, then update the task checkbox with [x]→Update story section File List to ensure it lists and new or modified or deleted source file→repeat order-of-execution until complete'
- story-file-updates-ONLY: - story-file-updates-ONLY:
- CRITICAL: ONLY UPDATE THE STORY FILE WITH UPDATES TO SECTIONS INDICATED BELOW. DO NOT MODIFY ANY OTHER SECTIONS. - CRITICAL: ONLY UPDATE THE STORY FILE WITH UPDATES TO SECTIONS INDICATED BELOW. DO NOT MODIFY ANY OTHER SECTIONS.
- CRITICAL: You are ONLY authorized to edit these specific sections of story files - Tasks / Subtasks Checkboxes, Dev Agent Record section and all its subsections, Agent Model Used, Debug Log References, Completion Notes List, File List, Change Log, Status - CRITICAL: You are ONLY authorized to edit these specific sections of story files - Tasks / Subtasks Checkboxes, Dev Agent Record section and all its subsections, Agent Model Used, Debug Log References, Completion Notes List, File List, Change Log, Status
- CRITICAL: DO NOT modify Status, Story, Acceptance Criteria, Dev Notes, Testing sections, or any other sections not listed above - CRITICAL: DO NOT modify Status, Story, Acceptance Criteria, Dev Notes, Testing sections, or any other sections not listed above
- blocking: 'HALT for: Unapproved deps needed, confirm with user | Ambiguous after story check | 3 failures attempting to implement or fix something repeatedly | Missing config | Failing regression' - blocking: 'HALT for: Unapproved deps needed, confirm with user | 3 failures attempting to implement or fix something repeatedly | Missing config | Failing regression'
- ready-for-review: 'Code matches requirements + All validations pass + Follows standards + File List complete' - ready-for-review: 'Code matches requirements + All validations pass + Follows standards + File List complete'
- completion: "All Tasks and Subtasks marked [x] and have tests→Validations and full regression passes (DON'T BE LAZY, EXECUTE ALL TESTS and CONFIRM)→Ensure File List is Complete→run the task execute-checklist for the checklist story-dod-checklist→set story status: 'Ready for Review'→HALT" - completion: "All Tasks and Subtasks marked [x] and have tests→Validations and full regression passes (DON'T BE LAZY, EXECUTE ALL TESTS and CONFIRM)→Ensure File List is Complete→run the task execute-checklist for the checklist story-dod-checklist→set story status: 'Ready for Review'→HALT"
# TDD-Specific Commands (only available when tdd.enabled=true)
- tdd-implement {story}: |
Execute tdd-implement task for TDD Green phase.
Implement minimal code to make failing tests pass. No feature creep.
Prerequisites: Story has failing tests (tdd.status='red'), test runner configured.
Outcome: All tests pass, story tdd.status='green', ready for refactor assessment.
- make-tests-pass {story}: |
Iterative command to run tests and implement fixes until all tests pass.
Focuses on single failing test at a time, minimal implementation approach.
Auto-runs tests after each change, provides fast feedback loop.
- tdd-refactor {story}: |
Collaborate with QA agent on TDD Refactor phase.
Improve code quality while keeping all tests green.
Prerequisites: All tests passing (tdd.status='green').
Outcome: Improved code quality, tests remain green, tdd.status='refactor' or 'done'.
# Utility Commands
- explain: teach me what and why you did whatever you just did in detail so I can learn. Explain to me as if you were training a junior engineer. - explain: teach me what and why you did whatever you just did in detail so I can learn. Explain to me as if you were training a junior engineer.
- review-qa: run task `apply-qa-fixes.md' - review-qa: run task `apply-qa-fixes.md'
- run-tests: Execute linting and tests - run-tests: Execute linting and tests
@ -74,8 +94,17 @@ commands:
dependencies: dependencies:
checklists: checklists:
- story-dod-checklist.md - story-dod-checklist.md
- tdd-dod-checklist.md
tasks: tasks:
- apply-qa-fixes.md - apply-qa-fixes.md
- execute-checklist.md - execute-checklist.md
- validate-next-story.md - validate-next-story.md
# TDD-specific tasks
- tdd-implement.md
- tdd-refactor.md
prompts:
- tdd-green.md
- tdd-refactor.md
config:
- test-runners.yaml
``` ```

37
bmad-core/agents/qa.md
View File

@ -37,8 +37,9 @@ agent:
icon: 🧪 icon: 🧪
whenToUse: | whenToUse: |
Use for comprehensive test architecture review, quality gate decisions, Use for comprehensive test architecture review, quality gate decisions,
and code improvement. Provides thorough analysis including requirements Test-Driven Development (TDD) test creation, and code improvement.
traceability, risk assessment, and test strategy. Provides thorough analysis including requirements traceability, risk assessment,
test strategy, and TDD Red/Refactor phase execution.
Advisory only - teams choose their quality bar. Advisory only - teams choose their quality bar.
customization: null customization: null
persona: persona:
@ -57,6 +58,10 @@ persona:
- Technical Debt Awareness - Identify and quantify debt with improvement suggestions - Technical Debt Awareness - Identify and quantify debt with improvement suggestions
- LLM Acceleration - Use LLMs to accelerate thorough yet focused analysis - LLM Acceleration - Use LLMs to accelerate thorough yet focused analysis
- Pragmatic Balance - Distinguish must-fix from nice-to-have improvements - Pragmatic Balance - Distinguish must-fix from nice-to-have improvements
- TDD Test-First - Write failing tests before any implementation (Red phase)
- Test Isolation - Ensure deterministic, fast, independent tests with proper mocking
- Minimal Test Scope - Focus on smallest testable behavior slice, avoid over-testing
- Refactoring Safety - Collaborate on safe code improvements while maintaining green tests
story-file-permissions: story-file-permissions:
- CRITICAL: When reviewing stories, you are ONLY authorized to update the "QA Results" section of story files - CRITICAL: When reviewing stories, you are ONLY authorized to update the "QA Results" section of story files
- CRITICAL: DO NOT modify any other sections including Status, Story, Acceptance Criteria, Tasks/Subtasks, Dev Notes, Testing, Dev Agent Record, Change Log, or any other sections - CRITICAL: DO NOT modify any other sections including Status, Story, Acceptance Criteria, Tasks/Subtasks, Dev Notes, Testing, Dev Agent Record, Change Log, or any other sections
@ -64,6 +69,7 @@ story-file-permissions:
# All commands require * prefix when used (e.g., *help) # All commands require * prefix when used (e.g., *help)
commands: commands:
- help: Show numbered list of the following commands to allow selection - help: Show numbered list of the following commands to allow selection
# Traditional QA Commands
- gate {story}: Execute qa-gate task to write/update quality gate decision in directory from qa.qaLocation/gates/ - gate {story}: Execute qa-gate task to write/update quality gate decision in directory from qa.qaLocation/gates/
- nfr-assess {story}: Execute nfr-assess task to validate non-functional requirements - nfr-assess {story}: Execute nfr-assess task to validate non-functional requirements
- review {story}: | - review {story}: |
@ -74,10 +80,26 @@ commands:
- risk-profile {story}: Execute risk-profile task to generate risk assessment matrix - risk-profile {story}: Execute risk-profile task to generate risk assessment matrix
- test-design {story}: Execute test-design task to create comprehensive test scenarios - test-design {story}: Execute test-design task to create comprehensive test scenarios
- trace {story}: Execute trace-requirements task to map requirements to tests using Given-When-Then - trace {story}: Execute trace-requirements task to map requirements to tests using Given-When-Then
# TDD-Specific Commands (only available when tdd.enabled=true)
- tdd-start {story}: |
Initialize TDD process for a story. Sets tdd.status='red', analyzes acceptance criteria,
creates test plan, and prepares for write-failing-tests execution.
Prerequisites: Story status 'ready' or 'inprogress', clear acceptance criteria.
- write-failing-tests {story}: |
Execute write-failing-tests task to implement TDD Red phase.
Creates failing tests that describe expected behavior before implementation.
Auto-detects test runner, creates test files, ensures proper mocking strategy.
Prerequisites: tdd-start completed or story ready for TDD.
- tdd-refactor {story}: |
Participate in TDD Refactor phase with Dev agent.
Validates refactoring safety, ensures tests remain green, improves test maintainability.
Collaborative command - works with Dev agent during refactor phase.
- exit: Say goodbye as the Test Architect, and then abandon inhabiting this persona - exit: Say goodbye as the Test Architect, and then abandon inhabiting this persona
dependencies: dependencies:
data: data:
- technical-preferences.md - technical-preferences.md
- test-levels-framework.md
- test-priorities-matrix.md
tasks: tasks:
- nfr-assess.md - nfr-assess.md
- qa-gate.md - qa-gate.md
@ -85,7 +107,18 @@ dependencies:
- risk-profile.md - risk-profile.md
- test-design.md - test-design.md
- trace-requirements.md - trace-requirements.md
# TDD-specific tasks
- write-failing-tests.md
- tdd-refactor.md
templates: templates:
- qa-gate-tmpl.yaml - qa-gate-tmpl.yaml
- story-tmpl.yaml - story-tmpl.yaml
- story-tdd-template.md
checklists:
- tdd-dod-checklist.md
prompts:
- tdd-red.md
- tdd-refactor.md
config:
- test-runners.yaml
``` ```

47
bmad-core/tasks/test-design.md
View File

@ -2,7 +2,7 @@
# test-design # test-design
Create comprehensive test scenarios with appropriate test level recommendations for story implementation. Create comprehensive test scenarios with appropriate test level recommendations for story implementation. Supports both traditional testing and Test-Driven Development (TDD) first approaches.
## Inputs ## Inputs
@ -12,12 +12,17 @@ required:
- story_path: '{devStoryLocation}/{epic}.{story}.*.md' # Path from core-config.yaml - story_path: '{devStoryLocation}/{epic}.{story}.*.md' # Path from core-config.yaml
- story_title: '{title}' # If missing, derive from story file H1 - story_title: '{title}' # If missing, derive from story file H1
- story_slug: '{slug}' # If missing, derive from title (lowercase, hyphenated) - story_slug: '{slug}' # If missing, derive from title (lowercase, hyphenated)
optional:
- tdd_mode: boolean # If true, design tests for TDD Red phase (before implementation)
- existing_tests: array # List of existing tests to consider for gap analysis
``` ```
## Purpose ## Purpose
Design a complete test strategy that identifies what to test, at which level (unit/integration/e2e), and why. This ensures efficient test coverage without redundancy while maintaining appropriate test boundaries. Design a complete test strategy that identifies what to test, at which level (unit/integration/e2e), and why. This ensures efficient test coverage without redundancy while maintaining appropriate test boundaries.
**TDD Mode**: When `tdd_mode=true`, design tests that will be written BEFORE implementation (Red phase), focusing on smallest testable behavior slices and proper mocking strategies.
## Dependencies ## Dependencies
```yaml ```yaml
@ -71,6 +76,46 @@ test_scenario:
description: 'What is being tested' description: 'What is being tested'
justification: 'Why this level was chosen' justification: 'Why this level was chosen'
mitigates_risks: ['RISK-001'] # If risk profile exists mitigates_risks: ['RISK-001'] # If risk profile exists
# TDD-specific fields (when tdd_mode=true)
tdd_phase: red|green|refactor # When this test should be written
mocking_strategy: mock|fake|stub|none # How to handle dependencies
test_data_approach: fixed|builder|random # How to generate test data
```
### 4a. TDD-Specific Test Design (when tdd_mode=true)
**Smallest-Next-Test Principle:**
- Design tests for the absolute smallest behavior increment
- Each test should drive a single, focused implementation change
- Avoid tests that require multiple features to pass
**Mocking Strategy Selection Matrix:**
| Dependency Type | Recommended Approach | Justification |
| --------------- | -------------------- | -------------------------------------- |
| External API | Mock | Control responses, avoid network calls |
| Database | Fake | In-memory implementation for speed |
| File System | Stub | Return fixed responses |
| Time/Date | Mock | Deterministic time control |
| Random Numbers | Stub | Predictable test outcomes |
| Other Services | Mock/Fake | Depends on complexity and speed needs |
**Test Data Strategy:**
```yaml
test_data_approaches:
fixed_data:
when: 'Simple, predictable scenarios'
example: "const userId = 'test-user-123'"
builder_pattern:
when: 'Complex objects with variations'
example: "new UserBuilder().withEmail('test@example.com').build()"
avoid_random:
why: 'Makes tests non-deterministic and hard to debug'
instead: 'Use meaningful, fixed test data'
``` ```
### 5. Validate Coverage ### 5. Validate Coverage

299
bmad-core/templates/tdd-quality-gates.md Normal file
View File

@ -0,0 +1,299 @@
<!-- Powered by BMAD™ Core -->
# TDD Quality Gates Template
Quality gate criteria and checkpoints for Test-Driven Development workflows.
## Gate Structure
Each TDD phase has specific quality gates that must be met before progression to the next phase.
## Red Phase Gates
### Prerequisites for Red Phase Entry
- [ ] Story has clear acceptance criteria
- [ ] Test runner detected or configured
- [ ] Story status is 'ready' or 'inprogress'
- [ ] TDD enabled in core-config.yaml
### Red Phase Completion Gates
**PASS Criteria:**
- [ ] At least one test written and failing
- [ ] Tests fail for correct reasons (missing implementation, not syntax errors)
- [ ] All external dependencies properly mocked
- [ ] Test data is deterministic (no random values, current time)
- [ ] Test names clearly describe expected behavior
- [ ] Story TDD metadata updated (status='red', test list populated)
- [ ] Test files follow project naming conventions
**FAIL Criteria:**
- [ ] No tests written
- [ ] Tests pass unexpectedly (implementation may already exist)
- [ ] Tests fail due to syntax errors or configuration issues
- [ ] External dependencies not mocked (network calls, file system, etc.)
- [ ] Non-deterministic tests (random data, time-dependent)
**Gate Decision:**
```yaml
red_phase_gate:
status: PASS|FAIL
failing_tests_count: { number }
tests_fail_correctly: true|false
mocking_complete: true|false
deterministic_tests: true|false
metadata_updated: true|false
ready_for_green_phase: true|false
```
## Green Phase Gates
### Prerequisites for Green Phase Entry
- [ ] Red phase gate passed
- [ ] Story tdd.status = 'red'
- [ ] Failing tests exist and documented
- [ ] Test runner confirmed working
### Green Phase Completion Gates
**PASS Criteria:**
- [ ] All previously failing tests now pass
- [ ] No new tests added during implementation
- [ ] Implementation is minimal (only what's needed for tests)
- [ ] No feature creep beyond test requirements
- [ ] All existing tests remain green (no regressions)
- [ ] Code follows basic quality standards
- [ ] Story TDD metadata updated (status='green')
**CONCERNS Criteria:**
- [ ] Implementation seems overly complex for test requirements
- [ ] Additional functionality added without corresponding tests
- [ ] Code quality significantly below project standards
- [ ] Performance implications not addressed
**FAIL Criteria:**
- [ ] Tests still failing after implementation attempt
- [ ] New regressions introduced (previously passing tests now fail)
- [ ] Implementation missing for some failing tests
- [ ] Significant feature creep detected
**Gate Decision:**
```yaml
green_phase_gate:
status: PASS|CONCERNS|FAIL
all_tests_passing: true|false
no_regressions: true|false
minimal_implementation: true|false
feature_creep_detected: false|true
code_quality_acceptable: true|false
ready_for_refactor_phase: true|false
```
## Refactor Phase Gates
### Prerequisites for Refactor Phase Entry
- [ ] Green phase gate passed
- [ ] Story tdd.status = 'green'
- [ ] All tests consistently passing
- [ ] Code quality issues identified
### Refactor Phase Completion Gates
**PASS Criteria:**
- [ ] All tests remain green throughout refactoring
- [ ] Code quality measurably improved
- [ ] No behavior changes introduced
- [ ] Refactoring changes committed incrementally
- [ ] Technical debt reduced in story scope
- [ ] Story TDD metadata updated (status='refactor' or 'done')
**CONCERNS Criteria:**
- [ ] Some code smells remain unaddressed
- [ ] Refactoring introduced minor complexity
- [ ] Test execution time increased significantly
- [ ] Marginal quality improvements
**FAIL Criteria:**
- [ ] Tests broken by refactoring changes
- [ ] Behavior changed during refactoring
- [ ] Code quality degraded
- [ ] Large, risky refactoring attempts
**Gate Decision:**
```yaml
refactor_phase_gate:
status: PASS|CONCERNS|FAIL
tests_remain_green: true|false
code_quality_improved: true|false
behavior_preserved: true|false
technical_debt_reduced: true|false
safe_incremental_changes: true|false
ready_for_completion: true|false
```
## Story Completion Gates
### TDD Story Completion Criteria
**Must Have:**
- [ ] All TDD phases completed (Red → Green → Refactor)
- [ ] Final test suite passes consistently
- [ ] Code quality meets project standards
- [ ] All acceptance criteria covered by tests
- [ ] TDD-specific DoD checklist completed
**Quality Metrics:**
- [ ] Test coverage meets story target
- [ ] No obvious code smells remain
- [ ] Test execution time reasonable (< 2x baseline)
- [ ] All TDD artifacts documented in story
**Documentation:**
- [ ] TDD cycle progression tracked in story
- [ ] Test-to-requirement traceability clear
- [ ] Refactoring decisions documented
- [ ] Lessons learned captured
## Gate Failure Recovery
### Red Phase Recovery
```yaml
red_phase_failures:
no_failing_tests:
action: 'Review acceptance criteria, create simpler test cases'
escalation: 'Consult SM for requirement clarification'
tests_pass_unexpectedly:
action: 'Check if implementation already exists, adjust test scope'
escalation: 'Review story scope with PO'
mocking_issues:
action: 'Review external dependencies, implement proper mocks'
escalation: 'Consult with Dev agent on architecture'
```
### Green Phase Recovery
```yaml
green_phase_failures:
tests_still_failing:
action: 'Break down implementation into smaller steps'
escalation: 'Review test expectations vs implementation approach'
regressions_introduced:
action: 'Revert changes, identify conflicting logic'
escalation: 'Architectural review with team'
feature_creep_detected:
action: 'Remove features not covered by tests'
escalation: 'Return to Red phase for additional tests'
```
### Refactor Phase Recovery
```yaml
refactor_phase_failures:
tests_broken:
action: 'Immediately revert breaking changes'
escalation: 'Use smaller refactoring steps'
behavior_changed:
action: 'Revert and analyze where behavior diverged'
escalation: 'Review refactoring approach with QA agent'
quality_degraded:
action: 'Revert changes, try different refactoring technique'
escalation: 'Accept current code quality, document technical debt'
```
## Quality Metrics Dashboard
### Per-Phase Metrics
```yaml
metrics_tracking:
red_phase:
- failing_tests_count
- test_creation_time
- mocking_complexity
green_phase:
- implementation_time
- lines_of_code_added
- test_pass_rate
refactor_phase:
- code_quality_delta
- test_execution_time_delta
- refactoring_safety_score
```
### Story-Level Metrics
```yaml
story_metrics:
total_tdd_cycle_time: '{hours}'
cycles_completed: '{count}'
test_to_code_ratio: '{percentage}'
coverage_achieved: '{percentage}'
quality_improvement_score: '{0-100}'
```
## Integration with Standard Gates
### How TDD Gates Extend Standard QA Gates
- **Standard gates still apply** for final story review
- **TDD gates are additional checkpoints** during development
- **Phase-specific criteria** supplement overall quality assessment
- **Traceability maintained** between TDD progress and story completion
### Gate Reporting
```yaml
gate_report_template:
story_id: '{epic}.{story}'
tdd_enabled: true
phases_completed: ['red', 'green', 'refactor']
phase_gates:
red:
status: 'PASS'
completed_date: '2025-01-01T10:00:00Z'
criteria_met: 6/6
green:
status: 'PASS'
completed_date: '2025-01-01T14:00:00Z'
criteria_met: 7/7
refactor:
status: 'PASS'
completed_date: '2025-01-01T16:00:00Z'
criteria_met: 6/6
final_assessment:
overall_gate: 'PASS'
quality_score: 92
recommendations: []
```
This template ensures consistent quality standards across all TDD phases while maintaining compatibility with existing BMAD quality gates.