What is 'cold-start' execution in Blueprint?

Cold-start execution means each step in a generated plan contains all the necessary context and exit criteria, allowing a fresh AI agent to execute the task without needing access to previous session history.

When should I use Blueprint instead of standard Claude Code commands?

Use Blueprint for tasks that require more than one PR, multiple work sessions, or complex dependency ordering. For simple tasks that take fewer than three tool calls, standard commands are more efficient.

Does Blueprint require the GitHub CLI to function?

No. While Blueprint detects Git and GitHub CLI to generate PR-based workflow plans, it automatically degrades to 'direct mode' for local file edits if those tools are not available.

How does the adversarial review gate improve plans?

Before finalizing a plan, Blueprint delegates a review to a high-tier sub-agent (like Claude 3.5 Sonnet or Opus) to check for dependency errors, missing steps, and common architectural anti-patterns.

Blueprint Plan Generator

Name: Blueprint Plan Generator
Author: saar120

bysaar120

0•

Productivity & Workflow

Transforms high-level engineering objectives into structured, multi-step construction plans optimized for multi-session AI execution.

Blueprint is a specialized workflow orchestration skill for Claude Code designed to solve context loss during complex, multi-day engineering projects. By breaking down one-line objectives into individual PR-sized steps, it creates a roadmap where every task includes its own self-contained context brief for 'cold-start' execution. The skill utilizes a sophisticated 5-phase pipeline—including research, dependency mapping, and an adversarial review gate—to ensure that large-scale migrations, refactors, and feature builds are logically sound and ready for parallel workstreams without the risk of architectural drift.

Key Features

01Self-contained context briefs for every plan step to enable cold-start execution.

02Adversarial review gate that uses high-tier models to catch anti-patterns and logic gaps.

03Plan mutation protocol for safely splitting, reordering, or updating active roadmaps.

04Adaptive workflow support that toggles between Git/GitHub PRs and direct-edit modes.

05Automatic dependency graph generation and parallel workstream detection.

060 GitHub stars

Use Cases

01Executing complex database migrations that require multiple sequential PRs and verification steps.

02Organizing parallel development of independent modules or plugins across different agent sessions.

03Mapping out large-scale architectural refactors to prevent context loss during long-term projects.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add saar120/skillgate-registry blueprint

For use in Claude.ai and ChatGPT

Key Features

01Self-contained context briefs for every plan step to enable cold-start execution.

02Adversarial review gate that uses high-tier models to catch anti-patterns and logic gaps.

03Plan mutation protocol for safely splitting, reordering, or updating active roadmaps.

04Adaptive workflow support that toggles between Git/GitHub PRs and direct-edit modes.

05Automatic dependency graph generation and parallel workstream detection.

060 GitHub stars

Use Cases

01Executing complex database migrations that require multiple sequential PRs and verification steps.

02Organizing parallel development of independent modules or plugins across different agent sessions.

03Mapping out large-scale architectural refactors to prevent context loss during long-term projects.