blueprint

Generate step-by-step engineering plans with dependency graphs and parallel step detection.

2|Updated Mar 12, 2026
One-click install
npx skills add https://github.com/sayasaya8039/ZWG_Terminal --skill blueprint-sayasaya8039
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: blueprint
Source: https://github.com/sayasaya8039/ZWG_Terminal/tree/main/.claude/skills/blueprint
Command: npx skills add https://github.com/sayasaya8039/ZWG_Terminal --skill blueprint-sayasaya8039

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the risk of context loss, rework, and misalignment when breaking large, complex engineering tasks (such as multi-PR feature development, cross-session refactors, or multi-agent coordinated work) into discrete, executable steps for fresh agents or team members.

Core Features & Use Cases

  • Adversarial Plan Review: Every generated plan is reviewed by a high-capability sub-agent against an anti-pattern catalog to catch gaps, dependency errors, and edge cases before execution.
  • Cold-Start Step Context: Each step includes a self-contained context brief, so any agent can execute it without reading prior steps or retaining prior session context.
  • Parallel Work Detection: Automatically identifies independent steps that can be run in parallel to speed up multi-session projects, with clear dependency graphs to avoid conflicts.
  • Use Case Example: Plan a 6-PR migration of a monolithic application to a microservices architecture, with parallel workstreams for different service domains and formal rollback strategies for each step.

Quick Start

Use the blueprint skill to create a step-by-step construction plan for adding OAuth 2.0 authentication to the existing web application.

Frequently Asked Questions about blueprint

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

FAQPage Schema
How do I break down a large engineering task into executable steps for multi-session execution?

To break down large engineering tasks for multi-session execution, generate self-contained step briefs with cold-start context. This ensures fresh agents can execute discrete steps without retaining prior session context, preventing rework and misalignment.

What is adversarial plan review for multi-PR feature development?

Adversarial plan review for multi-PR feature development is a validation mechanism where a high-capability sub-agent evaluates generated plans against an anti-pattern catalog. This process catches dependency errors, gaps, and edge cases before code execution begins.

How do I identify parallel workstreams and dependency graphs for cross-session refactors?

Identify parallel workstreams and dependency graphs for cross-session refactors by automatically detecting independent steps that can run concurrently. This parallel step detection speeds up multi-session projects while using clear dependency graphs to avoid execution conflicts.

Can I plan a monolith to microservices migration with parallel workstreams and rollback strategies?

Yes, you can plan a monolith to microservices migration with parallel workstreams and rollback strategies. This approach coordinates multiple PRs across different service domains while applying formal mutation protocols to ensure low-risk, executable project plans.

Why does context loss cause rework risks in multi-agent coordinated work?

Context loss causes rework risks in multi-agent coordinated work because fresh agents lack prior session memory, leading to misalignment. Self-contained step context briefs solve this by giving agents all necessary execution details without reading prior steps.

What are the limitations of using automated parallel step detection for complex migrations?

Automated parallel step detection for complex migrations requires formal mutation protocols and adversarial review to remain safe. Without high-capability sub-agent validation against anti-pattern catalogs, undetected dependency conflicts between independent steps can still occur.