How does this skill prevent context pollution?

It launches a fresh, isolated subagent for every task, providing only the specific files and CLAUDE.md sections relevant to that specific task's classification.

What happens if a task fails verification?

The skill will attempt to fix the issue and re-verify. If it remains blocked after three retries, it pauses execution and asks the user for a new approach or guidance.

Can I use this skill without a pre-existing plan?

No, it requires an approved plan file (located in docs/plans/) and a clean git state to ensure changes are tracked and validated correctly.

Does it support parallel task execution?

Yes, if the plan marks tasks as independent, the skill can launch multiple subagents simultaneously to accelerate the development process.

How does the skill handle decision logging?

Each subagent records non-trivial decisions regarding architecture, patterns, and trade-offs, which are then compiled into a structured execution trace for auditing.

Plan Execution Workflow

Name: Plan Execution Workflow
Author: Brite-Nites

byBrite-Nites

0•

生产力与工作流

Executes complex development plans using isolated subagents and strict TDD enforcement for high-quality code implementation.

The Plan Execution Workflow skill automates the implementation of structured software development plans by delegating tasks to specialized, isolated subagents. By maintaining strict context isolation—providing each subagent only the specific files and conventions needed for its assigned task—it prevents quality degradation caused by context noise. The skill ensures reliability through a rigorous Test-Driven Development (TDD) protocol, multi-level verification checkpoints, and detailed execution tracing, making it ideal for implementing large-scale features, complex refactors, and systematic bug fixes.

主要功能

01Strict TDD enforcement (Red-Green-Refactor) for every testable code change.

020 GitHub stars

03Four-level verification process covering builds, tests, acceptance criteria, and integration.

04Automated context selection that injects specific CLAUDE.md sections based on task classification.

05Parallel execution support for independent tasks to significantly reduce total implementation time.

06Subagent-per-task delegation to maintain clean context and prevent noise accumulation.

使用场景

01Executing large-scale refactoring efforts where systematic verification is critical.

02Implementing complex features that require changes across multiple architectural layers.

03Automating the transition from a design document to a fully tested implementation.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add brite-nites/brite-claude-plugins executing-plans

For use in Claude.ai and ChatGPT

主要功能

01Strict TDD enforcement (Red-Green-Refactor) for every testable code change.

020 GitHub stars

03Four-level verification process covering builds, tests, acceptance criteria, and integration.

04Automated context selection that injects specific CLAUDE.md sections based on task classification.

05Parallel execution support for independent tasks to significantly reduce total implementation time.

06Subagent-per-task delegation to maintain clean context and prevent noise accumulation.

使用场景

01Executing large-scale refactoring efforts where systematic verification is critical.

02Implementing complex features that require changes across multiple architectural layers.

03Automating the transition from a design document to a fully tested implementation.