execute-task

name	execute-task
description	Execute approved task specifications sequentially with TDD, comprehensive testing, and validation. This skill should be used for implementing tasks from approved specs with full audit trail.
acceptance	[{"task_loaded":"Task specification successfully loaded and validated"},{"status_updated":"Task status progressed from Approved → InProgress → Review"},{"all_complete":"All tasks and subtasks marked complete"},{"tests_passing":"All tests written and passing"},{"criteria_verified":"All acceptance criteria verified and documented"},{"record_complete":"Implementation Record fully populated"}]
inputs	{"task_file":{"type":"string","required":true,"description":"Path to task specification file (e.g., .claude/tasks/task-006.md)","validation":"Must be existing file with status 'Approved'"},"auto_confirm":{"type":"boolean","required":false,"description":"Skip execution plan confirmation","default":false}}
outputs	{"implementation_complete":{"type":"boolean","description":"Whether all tasks completed successfully"},"tasks_completed":{"type":"number","description":"Number of tasks completed"},"subtasks_completed":{"type":"number","description":"Number of subtasks completed"},"tests_passed":{"type":"boolean","description":"Whether all tests passed"},"total_tests":{"type":"number","description":"Total number of tests executed"},"files_modified":{"type":"array","description":"List of files created or modified"},"status":{"type":"string","description":"Final task status (should be 'Review')"}}
telemetry	{"emit":"skill.execute-task.completed","track":["task_file","tasks_completed","subtasks_completed","tests_passed","total_tests","files_modified_count","duration_ms","halt_count"]}

Execute Task Skill

Purpose

Execute approved task specifications sequentially using Test-Driven Development (TDD), with comprehensive validation at each step and full implementation audit trail.

Core Capabilities:

• Sequential task/subtask execution with validation gates
• Permission-controlled file modifications
• Comprehensive testing and documentation
• Implementation Record maintenance
• Halt-on-error safeguards

BMAD Pattern (Key Innovation):

• ALL context embedded in task spec by planning skill
• Dev reads ONLY task spec + always-load files (coding standards)
• No architecture lookup during implementation
• Result: Focused execution, no context searching, no drift

Prerequisites

• Task specification file with status "Approved"
• Configuration file (.claude/config.yaml) with development settings
• Always-load files (coding standards) available
• Test framework configured (Jest, Pytest, etc.)

---

Workflow

Step 0: Load Configuration and Task Specification

Action: Load all required context and verify task readiness.

Execute:

python .claude/skills/bmad-commands/scripts/read_file.py \
  --path .claude/config.yaml \
  --output json

python .claude/skills/bmad-commands/scripts/read_file.py \
  --path {task_file} \
  --output json

Parse Response:

• Extract development.alwaysLoadFiles from config
• Extract task status, objective, acceptance criteria, tasks
• Verify status is "Approved"

Validation:

• Configuration file exists and loadable
• Task file exists and loadable
• Task status is "Approved" (halt if Draft/Review/Done)
• Always-load files exist

Update Status:

• Change task status from "Approved" to "InProgress"
• Record start time in Implementation Record

See: references/configuration-guide.md for detailed configuration loading

---

Step 1: Review Task Context and Plan Execution

Action: Review embedded context and present execution plan to user.

Context Review: