autoplan

Orchestrate CEO, design, and engineering skill reviews with auto-decisions.

Updated Apr 3, 2026
One-click install
npx skills add https://github.com/lyh2/gstack-deepseek --skill autoplan-lyh2
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: autoplan
Source: https://github.com/lyh2/gstack-deepseek/tree/main/autoplan
Command: npx skills add https://github.com/lyh2/gstack-deepseek --skill autoplan-lyh2

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Auto-review pipeline orchestrates a full run of CEO, design, and Eng reviews from disk and applies auto-decisions to deliver a fully vetted plan.

Core Features & Use Cases

  • End-to-end plan review orchestration across CEO, design, and engineering skills with automatic decisioning.
  • Taste-level notes surfaced at a final gate to help human reviewers understand edge cases.
  • Single-command operation: /autoplan to produce a complete plan from the loaded skill set.

Quick Start

Run /autoplan to generate a fully reviewed plan from the loaded skill files.

Frequently Asked Questions about autoplan

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate end-to-end plan review for engineering, design, and CEO gates?

Automated plan review orchestrates CEO, design, and engineering skills from disk to apply auto-decisions and deliver a fully vetted plan. Running a single command produces a complete, ready-to-ship plan with minimal human input.

What is an auto-decision pipeline for plan gates and how does it work?

An auto-decision pipeline reads loaded skills from disk and applies six decision principles to drive approvals across CEO, design, and engineering reviews. It surfaces taste-level notes at a final gate to flag edge cases for human reviewers.

Do I need existing skill files on disk to run an automated plan review?

Yes, automated plan review requires CEO, design, and engineering skills loaded on disk with the required frontmatter in SKILL.md. It uses these existing skill files to orchestrate the full review sequence and apply automatic decisions.

What is the best way to produce a ready-to-ship plan with minimal manual input?

The best way to produce a ready-to-ship plan with minimal manual input is running a single orchestration command. It sequentially applies CEO, design, and engineering reviews, using six decision principles to automate approvals and surface final taste-level notes.

How do I surface taste-level notes and edge cases during a plan review?

Surfacing taste-level notes during plan review happens at the final gate after automated decisions are applied. The pipeline flags edge cases from the CEO, design, and engineering reviews to help human reviewers understand nuanced context before shipping.

What are the limitations of fully automating design and engineering plan reviews?

Fully automating design and engineering plan reviews limits human input to a final gate where taste-level notes are surfaced. While auto-decisions drive the pipeline using six principles, nuanced edge cases still require human review before the plan ships.