onboard

Migrates existing project documentation into framework gate artifacts through semantic content matching.

Updated Apr 13, 2026
One-click install
npx skills add https://github.com/scoots31/engineering-playbook --skill onboard-scoots31
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: onboard
Source: https://github.com/scoots31/engineering-playbook/tree/main/skills/onboard
Command: npx skills add https://github.com/scoots31/engineering-playbook --skill onboard-scoots31

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Projects started outside the framework have decisions and plans scattered across READMEs, notes, and docs in non-standard formats. This Skill translates that existing work into framework-format gate files without redoing completed phases, and identifies the correct starting phase. ## Core Features & Use Cases - Semantic Content Mapping: Scans the project directory and matches existing content to framework gate artifacts (discovery brief, tech context, backlog, process maps) by meaning, not filename. - Gap Classification: Surfaces missing artifacts as targeted discussion questions, then classifies remaining gaps as undocumented work or work not yet done to determine the starting phase. - Continuity Initialization: Creates current-phase.md, handoff.md, and decisions.md so the first real session starts with full context. - Use Case: You have a half-planned app with notes in a README and a tasks file. Run onboard to convert them into a discovery brief and backlog, then get placed at the Design Review phase automatically. ## Quick Start Tell the assistant: I have prior work on my project at /path/to/project and want to bring it into the framework — onboard it.

Frequently Asked Questions about onboard

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

FAQPage Schema
How do I bring an existing project into a structured development framework?

Point the onboard skill at your project directory and it scans markdown files, notes, and docs for content matching framework gate artifacts. It shows each proposed mapping for confirmation before writing any file, so existing work is translated rather than redone.

How does project onboarding determine which phase to start at?

After mapping existing content, remaining gaps are classified as undocumented work (asked again in conversation) or work not done yet. The first phase with genuinely missing work becomes the starting phase, recorded in docs/continuity/current-phase.md.

Does onboarding match files by name or by content?

Matching is semantic, based on content rather than filenames. A file named ff-eval-notes.md can map to the discovery brief if it contains use cases and problem statements, and content spread across multiple files gets synthesized into one artifact.

What files does onboarding skip when scanning a project?

It skips code files, build artifacts, node_modules, .git, dist, virtual environments, binaries, images, and lock files. Code is only read if no documentation exists at all, keeping the scan focused on high-signal planning content.

What happens if I cannot answer a gap question during onboarding?

The gap is classified as work not done yet, which defines your starting phase, and onboarding moves on. The loop never blocks indefinitely on unanswered questions.