What is the primary purpose of the do-task skill?

The do-task skill provides a structured workflow to execute coding tasks by moving through planning, implementation, and evaluation phases automatically.

Where does the skill store task documentation?

All task requests, plans, implementation summaries, and evaluation reports are stored in a versioned folder within the 'task-results/' directory.

Does it support automated testing?

Yes, the evaluation phase requires a runnable verification script to ensure the implementation actually works, rather than just performing a static code review.

What subagents are required for this skill?

This skill orchestrates three specific subagents: dt-task-planner, dt-task-implementer, and dt-task-evaluator.

How does the retry logic function?

If the evaluation phase returns a FAIL result, the skill increments an attempt counter and sends the failure feedback back to the implementer subagent to correct the issues.

Task Execution Workflow Runner

Name: Task Execution Workflow Runner
Author: ETdoFresh

byETdoFresh

•

Productivity & Workflow

Automates complex coding tasks through a structured plan-implement-evaluate lifecycle with integrated retry logic.

The do-task skill is an advanced orchestration framework for Claude Code that manages the end-to-end lifecycle of software development tasks. By coordinating specialized subagents for planning, implementation, and evaluation, it ensures that every task is documented, executed against a specific plan, and verified with runnable tests. It automatically handles errors and retries failed implementations by feeding evaluation feedback back into the loop, making it an essential tool for reliable, autonomous software engineering and complex refactoring projects.

Key Features

01Automatic retry logic that loops until verification tests pass successfully

02Structured directory management with persistent logging of every task phase

03Mandatory runnable verification scripts to ensure code quality and functional correctness

041 GitHub stars

05Multi-phase orchestration involving planning, implementation, and evaluation agents

06Automated task slug generation and documentation for better project traceability

Use Cases

01Complex codebase refactoring requiring multi-step verification and error correction

02Autonomous feature implementation based on high-level task descriptions

03Automated bug fixing where the fix must be validated by a test script before completion

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add etdofresh/claude-marketplace do-task

For use in Claude.ai and ChatGPT

Download Skill