Execute the implementation plan by processing and executing all tasks defined in tasks.md
Executes a complete implementation plan from tasks.md, running phases sequentially with TDD and validation. Use this to implement all planned tasks for a feature in one go.
/plugin marketplace add NeoLabHQ/context-engineering-kit/plugin install sdd@context-engineering-kitOptional implementation preferences or specific tasks to prioritize$ARGUMENTS
You MUST consider the user input before proceeding (if not empty).
feature/<number-padded-to-3-digits>-<kebab-case-title>, part after feature/ is defined as FEATURE_NAME. Consuquently, FEATURE_DIR is defined as specs/FEATURE_NAME.Goal: Implement taks list written in FEATURE_DIR/tasks.md file.
Actions:
For each phase in tasks.md file perform following actions:
Execute implementation by launching new developer agent to implement each phase, verify that all tasks are completed in order and without errors:
Use provided prompt exactly, while prefiling required variables:
**Goal**: Implement {phase name} phase of tasks.md file by following Tasks.md Execution Workflow.
User Input: {provide user input here if it exists}
FEATURE_NAME: {FEATURE_NAME}
FEATURE_DIR: {FEATURE_DIR}
TASKS_FILE: {TASKS_FILE}
Progress tracking and error handling:
Completion validation - Launch new developer agent to verify that all tasks are completed in order and without errors by using provided prompt exactly, while prefiling required variables:
**Goal**: Verify that all tasks in tasks.md file are completed in order and without errors.
- Verify all required tasks are completed
- Check that implemented features match the original specification
- Validate that tests pass and coverage meets requirements
- Confirm the implementation follows the technical plan
- Report final status with summary of completed work
User Input: {provide user input here if it exists}
FEATURE_NAME: {FEATURE_NAME}
FEATURE_DIR: {FEATURE_DIR}
TASKS_FILE: {TASKS_FILE}
If not all phases are completed, repeat steps 1-4 for the next phase.
Perform /code-review:review-local-changes command if it is available, if not then launch 3 developer agent to review code quality by using provided prompt exactly, while prefiling required variables, each of them should focus on different aspect of code quality: simplicity/DRY/elegance, bugs/functional correctness, project conventions/abstractions. Prompt for each agent:
**Goal**: Tasks.md file is implemented, review newly implemented code. Focus on {focus area}.
User Input: {provide user input here if it exists}
FEATURE_NAME: {FEATURE_NAME}
FEATURE_DIR: {FEATURE_DIR}
TASKS_FILE: {TASKS_FILE}
Consolidate findings and identify highest severity issues that you recommend fixing
Present findings to user and ask what they want to do (fix now, fix later, or proceed as-is)
Launch new developer agent to address issues based on user decision