Executing Plans

This is designed for when you've already written out an implementation plan and want Claude to execute it methodically with review checkpoints. It breaks work into batches (default is three tasks at a time), reports back after each batch, and waits for your feedback before continuing. The critical bit is the upfront plan review where Claude is supposed to flag concerns before starting rather than improvising mid-execution. It explicitly tells Claude to stop and ask when blocked instead of guessing, which is the right instinct but depends heavily on having a well-structured plan to begin with. Chains into the finishing-a-development-branch skill when tasks are complete. Most useful when you want tight control over implementation pacing.

skills.sh installs: 586
GitHub stars: 43.1K
last updated: Jul 14, 2026

View on GitHub

Install

$ npx -y skills add sickn33/antigravity-awesome-skills --skill executing-plans --agent claude-code

Installs into .claude/skills of the current project.

Files

Executing Plans

Overview

Load plan, review critically, execute tasks in batches, report for review between batches.

Core principle: Batch execution with checkpoints for architect review.

Announce at start: "I'm using the executing-plans skill to implement this plan."

The Process

Step 1: Load and Review Plan

Read plan file
Review critically - identify any questions or concerns about the plan
If concerns: Raise them with your human partner before starting
If no concerns: use the project’s existing task/todo mechanism and proceed; TodoWrite is a host-specific tool, not a universally installed command

Step 2: Execute Batch

Batch size follows the user’s explicit instruction or project contract; otherwise choose a practical reviewable group

For each task:

Mark as in_progress
Follow each step exactly (plan has bite-sized steps)
Run verifications as specified
Mark as completed

Step 3: Report

When batch complete:

Show what was implemented
Show verification output
Continue to the next authorized batch unless the user explicitly requested a checkpoint; provide concise progress without ending the task unnecessarily.

SKILL.md on GitHub