blueprint

Generate a step-by-step construction plan from a single objective.

2|Updated Mar 21, 2026
One-click install
npx skills add https://github.com/Throokie/claude-code-skills --skill blueprint-throokie
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: blueprint
Source: https://github.com/Throokie/claude-code-skills/tree/main/skills/blueprint
Command: npx skills add https://github.com/Throokie/claude-code-skills --skill blueprint-throokie

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Turn a one-line objective into a structured, executable plan that can be carried out by autonomous agents across multiple sessions, preserving context and enabling reliable coordination.

Core Features & Use Cases

  • Self-contained step contexts so new agents can execute steps without prior memory
  • Dependency graph and parallel-step detection to optimize sequencing
  • Adversarial review gate to ensure completeness, correctness, and safety
  • Multi-agent collaboration support across sessions and tools

Quick Start

Provide a project objective to generate a full blueprint with steps, owners, and validation checks.

Frequently Asked Questions about blueprint

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

FAQPage Schema
How do I break large engineering features into PR-sized steps for autonomous agents?

To break large features into PR-sized steps, you can generate a structured multi-session execution plan with self-contained contexts. This enforces a dependency graph and per-step context for cold-start execution by autonomous agents.

What is a dependency graph for multi-agent planning pipelines?

A dependency graph for multi-agent planning pipelines maps task sequencing and detects parallel steps. It optimizes sub-agent coordination by ensuring independent workstreams execute concurrently while preserving execution order.

How do I preserve context for multi-session software engineering projects?

You preserve context for multi-session projects by generating self-contained step contexts within the execution plan. This allows new agents to execute tasks without prior memory, ensuring reliable coordination across different tools.

Does the adversarial review gate check plan completeness before execution?

Yes, the adversarial review gate checks plan completeness, correctness, and safety before execution. It serves as a validation phase in the planning pipeline to ensure the generated construction steps are reliable for autonomous agents.

Can I use this to coordinate parallel workstreams across sub-agents?

Yes, you can coordinate parallel workstreams across sub-agents by leveraging the dependency graph and parallel-step detection. This optimizes sequencing and enables multi-agent collaboration across different sessions.