Does this skill require Test-Driven Development?

Yes, TDD is mandatory by default. This ensures every task is verified by tests before completion, unless the task is explicitly exempt, such as documentation or configuration updates.

How do I start the execution process?

The skill is typically triggered after using 'harness:writing-plans'. You announce the use of the skill and choose between Subagent or Batch mode based on your session requirements.

What is the difference between Subagent and Batch mode?

Subagent mode operates within the same session using isolated processes for faster iteration, while Batch mode is designed for separate sessions and complex changes requiring human feedback between tasks.

Can I use this skill for refactoring?

Absolutely. It is highly effective for refactoring as it manages task context isolation and ensures that existing or new tests pass at every stage of the execution plan.

How does it handle code quality?

The skill utilizes a multi-stage review process involving specialized subagents for both specification compliance and code quality audits, ensuring the output meets project standards.

Plan Execution & TDD Runner

Name: Plan Execution & TDD Runner
Author: astrosteveo

byastrosteveo

生産性とワークフロー

Executes complex software implementation plans through automated task-by-task workflows and mandatory test-driven development.

概要

The plan-execution skill provides a rigorous framework for transforming written implementation plans into production-ready code. By utilizing two distinct modes—Subagent for rapid, context-isolated execution and Batch for human-gated checkpoints—it ensures that complex changes are handled systematically. The skill enforces high engineering standards by mandating Test-Driven Development (TDD) and multi-stage review gates for specification compliance and code quality, effectively preventing context pollution and reducing technical debt during the development process.

主な機能

Automated review gates for specification compliance and code quality standards.
Seamless integration with the Harness agentic skills framework for end-to-end development.
0 GitHub stars
Mandatory Test-Driven Development (TDD) workflow for verified code output.
Context-isolated task execution to prevent LLM drift and hallucination.
Dual-mode execution (Subagent and Batch) to balance speed and human oversight.

ユースケース

Implementing complex features that require multi-file changes and strict architectural alignment.
Managing high-stakes development projects that require consistent code quality reviews.
Executing large-scale refactors where automated testing and task tracking are critical.

概要

主な機能

Automated review gates for specification compliance and code quality standards.
Seamless integration with the Harness agentic skills framework for end-to-end development.
0 GitHub stars
Mandatory Test-Driven Development (TDD) workflow for verified code output.
Context-isolated task execution to prevent LLM drift and hallucination.
Dual-mode execution (Subagent and Batch) to balance speed and human oversight.

ユースケース

Implementing complex features that require multi-file changes and strict architectural alignment.
Managing high-stakes development projects that require consistent code quality reviews.
Executing large-scale refactors where automated testing and task tracking are critical.