/speckit.implement command executes your implementation plan by systematically processing and completing all tasks defined in tasks.md.
Purpose
Automate feature implementation:- Execute tasks in correct dependency order
- Run parallel tasks concurrently where possible
- Follow Test-Driven Development (if tests included)
- Create/verify ignore files for the project
- Track progress and handle errors
- Mark completed tasks in tasks.md
- Validate implementation against specification
Implementation follows a phase-by-phase approach: Setup → Foundational → User Stories (by priority) → Polish. Each phase must complete before the next begins.
Usage
How It Works
Run Prerequisites
Executes prerequisite check script:
- Validates
FEATURE_DIRexists - Confirms
tasks.mdis present (errors if missing) - Parses
AVAILABLE_DOCSlist - Loads absolute paths for all artifacts
Check Checklists (if any)
Scans
checklists/ directory (if exists):- Counts total, completed, and incomplete items per checklist
- Creates status table
- If any checklist incomplete: STOPS and asks user to confirm proceeding
- If all complete: automatically proceeds
Load Implementation Context
Reads required and optional documents:
- Required: tasks.md (task list), plan.md (architecture)
- If exists: data-model.md, contracts/, research.md, quickstart.md
Project Setup Verification
Creates/verifies ignore files based on detected technologies:
- Checks if git repo: creates/verifies
.gitignore - Checks for Docker: creates/verifies
.dockerignore - Checks for ESLint: creates/verifies
.eslintignoreoreslint.config.*ignores - Checks for Prettier: creates/verifies
.prettierignore - Adds technology-specific patterns from plan.md tech stack
Parse Task Structure
Extracts from tasks.md:
- Task phases (Setup, Foundational, User Stories, Polish)
- Task IDs, descriptions, file paths
- Parallel markers
[P] - User story labels
[US1],[US2], etc. - Dependency relationships
Execute Phase-by-Phase
Processes each phase in order:
- Setup: Project initialization
- Foundational: Blocking prerequisites
- User Story phases: Independent feature slices (by priority)
- Polish: Cross-cutting improvements
- Respects dependencies (sequential vs parallel)
- Follows TDD (tests before implementation, if included)
- Validates completion before moving on
Progress Tracking
Reports after each task:
- Task completed
- Files created/modified
- Marks task as
[X]in tasks.md - Errors and suggested fixes (if task fails)
Completion Validation
Verifies:
- All required tasks completed
- Implementation matches specification
- Tests pass (if included)
- Technical plan followed
Checklist Validation
Before implementation starts, validates any checklists:Project Setup: Ignore Files
Before implementation, creates/verifies ignore files:Detection Logic
Technology-Specific Patterns
Based on tech stack from plan.md:Execution Flow
Phase 1: Setup
Phase 2: Foundational
Phase 3+: User Stories
Final Phase: Polish
Progress Tracking
After each task, marks completion in tasks.md:Error Handling
Non-Parallel Task Failure
Parallel Task Failure
Completion Validation
After all tasks:Real-World Example
Best Practices
Review Before Starting
Before running implement:- Review tasks.md to understand what will be built
- Complete all checklists (or consciously decide to proceed without)
- Ensure development environment is ready
Run Analyze First (Optional)
Let It Run
The command handles:- Dependency ordering
- Parallel execution
- File creation
- Test execution (if included)
- Progress tracking
Commit After Completion
Handoffs
After implementation:Manual Testing
Test the implemented features manually
Create Pull Request
Open PR for code review and merge
Deploy
Deploy to staging/production environment
Demo
Demonstrate completed user stories to stakeholders
File Updates
Next Steps
Create Pull Request
Review implementation and open PR for merge to main branch