skills.rest
michaeldyrynda
michaeldyryndaCOMMUNITY
Explore all skills by this author →
Website𝕏@michaeldyryndaGitHub

task-implementor

Implement one planned task using test-first development and pause for review.

59|3|Updated Jan 23, 2020
Search Tags:#automation#planning#handoff#plan#task-management#test-driven-development#vertical-slice
One-click install
npx skills add https://github.com/michaeldyrynda/dotfiles --skill task-implementor
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: task-implementor
Source: https://github.com/michaeldyrynda/dotfiles/tree/main/agents/skills/task-implementor
Command: npx skills add https://github.com/michaeldyrynda/dotfiles --skill task-implementor

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Implements exactly one planned task from .ai/tasks/ using test-first development, then pauses for human review. Invoke with /task-implementor.

Core Features & Use Cases

  • Executes exactly one task per invocation to enforce focus and deterministic progress.
  • Reads plan specs, learnings, and task metadata from the plan directory at .ai/plans/{slug}/.
  • Updates the chosen task's status to in_progress and surfaces a structured handoff summary after completion.

Quick Start

Invoke this skill to implement exactly one planned task from the current plan and pause for human review.

Frequently Asked Questions about task-implementor

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I execute a single planned task using test-driven development?▼

The task implementor picks one qualifying task from your project plan, applies a red-green-refactor TDD cycle, marks the task in_progress, and outputs a structured handoff summary for human review.

What is a test-driven handoff in vertical slice engineering?▼

A test-driven handoff is a structured summary generated after executing exactly one planned task via red-green-refactor TDD, ensuring deterministic progress and pausing for human review before proceeding.

How do I manage task execution from plan specs stored in my project directory?▼

The task implementor reads plan specs, learnings, and task metadata from .ai/plans/{slug}/, executes one task per invocation, and updates its status to in_progress for deterministic progress.

Can I process multiple tasks at once with test-first development automation?▼

No, this approach executes exactly one planned task per invocation to enforce focus and deterministic progress, requiring a human review pause before the next task is picked.

Do I need existing plan specs and task metadata before using task automation?▼

Yes, you need existing plan specs, learnings, and task metadata stored under .ai/plans/{slug}/, as the implementor reads these files to pick and execute a qualifying task.

What's the best way to ensure deterministic progress during plan-driven engineering?▼

Executing exactly one task per invocation using test-first development and pausing for human review with a detailed handoff summary ensures deterministic progress during plan-driven engineering.

Related Skills

Explore More Software Engineering Skills→
HonigbartHonigbart
1

quick-critique

Critique final_plan.md files with Codex CLI and write critique reports.

Community
Intermediate
timequitytimequity
6

verification-before-completion

Enforce a verification gate before claiming task completion.

Official
Basic
KartikSharma4448KartikSharma4448

project_tasks

Creates and manages tracked project tasks and milestones with lifecycle states.

Community
Intermediate
nexus-substratenexus-substrate
18

codex-delegator

Delegate code generation tasks to the Codex CLI for implementation and test writing.

Official
Intermediate