Can I use this for existing projects?

Yes, as long as you have an implementation plan (ideally created via the brainstorm skill), this skill can manage the execution of that plan on any codebase.

How does this skill improve AI coding reliability?

It enforces a disciplined workflow that requires real-world verification at every step, preventing the AI from skipping tests or making too many changes at once.

What happens if a test fails during execution?

The workflow is designed to stop immediately. This prevents Claude from pushing through blockers or layering new code on top of broken functionality.

Does this skill automatically commit my code?

It uses the /feature-tree:commit command after each successful testing phase to ensure your git history remains atomic and descriptive.

What is 'Real Testing' in the context of this skill?

Real Testing means using actual browser automation, live API calls, and real database queries instead of relying solely on unit tests or mocked data.

Plan Execution & Commit Workflow

Name: Plan Execution & Commit Workflow
Author: Nothflare

byNothflare

•

Productivity & Workflow

Executes complex implementation plans through a disciplined loop of incremental coding, real-world testing, and atomic commits.

The Plan Execution skill transforms abstract implementation plans into concrete code by enforcing a strict, step-by-step development cycle. It guides Claude through a structured loop: implementing specific feature groups, performing 'real' (non-mocked) testing, and committing changes via the feature-tree system. This skill is specifically designed to prevent the AI from becoming overwhelmed by large tasks, ensuring that every change is verified against actual environments—such as real databases or browser automation—before proceeding, thereby maintaining a clean and manageable project state.

Key Features

01Enforces a structured 'Implement-Test-Commit' loop for reliable development.

02Handles complex dependencies by following predefined implementation orders.

03Integrates with feature-tree commits to maintain a clean git history and symbol map.

04Automates progress reporting and state management between commit groups.

05Mandates real-world testing over mocks to prevent AI hallucinations and regressions.

061 GitHub stars

Use Cases

01Managing large-scale refactoring projects that require incremental verification.

02Executing multi-stage feature rollouts defined during a brainstorming session.

03Ensuring high-fidelity testing in web, API, and database-heavy environments.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add nothflare/feature-tree executing-plans

For use in Claude.ai and ChatGPT

Download Skill