How does this skill save context window space?

It uses a just-in-time loading strategy, reading only the header of the plan initially and loading specific phase details only when they are being actively executed.

What happens if a plan has missing tests?

The skill identifies functionality tasks without corresponding tests as 'plan gaps' and will pause implementation to ensure quality standards are met before proceeding.

Can I monitor the work of the subagents?

Yes, transparency is mandatory. The skill is required to print the full, unsummarized response of every subagent (implementor, fixer, or reviewer) to the user.

Is it safe to run multiple execution sessions?

Yes, the skill creates session-isolated scratchpad directories to prevent file collisions when multiple implementation plans are running simultaneously.

Does it automatically fix code review issues?

Yes, it dispatches a specialized task-bug-fixer subagent to address all issues identified during the review phase, including minor suggestions, until zero issues remain.

Implementation Plan Executor

Name: Implementation Plan Executor
Author: ed3dai

byed3dai

•

生产力与工作流

Executes complex implementation plans phase-by-phase using specialized subagents and rigorous review loops.

This skill provides a structured framework for executing multi-phase implementation plans within Claude Code. It optimizes performance by loading phase details 'just-in-time' to minimize context window usage and ensures high-quality output by dispatching dedicated subagents for implementation and bug fixing. With a heavy emphasis on human transparency, the skill keeps users informed of every subagent action while enforcing mandatory code review cycles and test coverage verification to turn architectural blueprints into production-grade code.

主要功能

01Session-isolated scratchpads for safe parallel execution

02Automated subagent dispatching for implementation and bug fixing

03Mandatory human transparency with full subagent response visibility

04Integrated multi-phase code review loops and issue tracking

05Just-in-time phase loading to optimize context window usage

065 GitHub stars

使用场景

01Executing complex architectural migrations or large-scale refactors

02Systematically implementing features from pre-defined technical specifications

03Ensuring rigorous coding standards and test coverage across multi-task projects

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ed3dai/ed3d-plugins-testing executing-an-implementation-plan

For use in Claude.ai and ChatGPT

Download Skill