Can I adjust the number of tasks in a batch?

Yes. While the default is set to three tasks per batch to ensure frequent checkpoints, you can provide feedback to the AI to adjust the batch size based on the complexity of the work.

Why is the initial review step necessary?

The initial review allows the AI to identify potential architectural flaws or missing information in the plan before any code is written, saving time and preventing rework.

How does this skill handle errors during execution?

The skill follows a strict stop-and-ask protocol. It pauses execution immediately if a test fails, a dependency is missing, or an instruction becomes unclear, preventing the AI from guessing.

What happens once the entire plan is finished?

The skill includes a required handoff to a finalization sub-skill, which performs comprehensive testing and prepares the development branch for completion.

Structured Plan Execution

Name: Structured Plan Execution
Author: Cycleaddict

byCycleaddict

0•

Productividad y Flujo de Trabajo

Implements complex development plans through a systematic, batch-oriented execution process with integrated review checkpoints.

Structured Plan Execution is a specialized skill for Claude Code designed to bridge the gap between architectural planning and hands-on implementation. By breaking down large implementation plans into manageable batches—typically three tasks at a time—the skill ensures that Claude remains aligned with human oversight through mandatory reporting and feedback loops. It emphasizes a critical pre-execution review, strict adherence to planned steps, and consistent verification, preventing 'hallucination creep' and unverified changes in complex development environments. This skill is essential for developers who want to maintain high quality and control during multi-step AI-assisted coding sessions.

Características Principales

010 GitHub stars

02Mandatory verification steps for every completed task to ensure code integrity.

03Batch-oriented task execution to maintain focus and allow for frequent human feedback.

04Strict 'stop-and-ask' protocol for blockers, dependencies, or unclear instructions.

05Integrated sub-skill handoff for final development branch completion and testing.

06Pre-execution critical review phase to identify plan gaps or blockers before coding begins.

Casos de Uso

01Implementing complex feature requests from a detailed technical specification.

02Executing multi-step refactorings where verification at each stage is critical.

03Systematic bug fixing across multiple files based on a pre-defined remediation plan.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add cycleaddict/generic-superpowers executing-plans

For use in Claude.ai and ChatGPT

Download Skill