bootstrap

Scaffold an empty decided repo and verify pipeline preconditions with a checked report.

2|Updated Jul 24, 2026
One-click install
npx skills add https://github.com/konstantinos-malavazos/claude-code-playbook --skill bootstrap-konstantinos-malavazos
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bootstrap
Source: https://github.com/konstantinos-malavazos/claude-code-playbook/tree/main/templates/skills/bootstrap
Command: npx skills add https://github.com/konstantinos-malavazos/claude-code-playbook --skill bootstrap-konstantinos-malavazos

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? After charting has decided the stack, layer chain, and Serena verdict, the repo is still empty. This Skill turns those decisions into a real, verified repository — stub, CLAUDE.md, Serena index, layer specialists, tracker adapter check, and two memories — without making any new decisions, and reports pass/fail on every seam check. ## Core Features & Use Cases - Eight-step scaffolding pipeline: asks allowlist questions, scaffolds the stub, writes CLAUDE.md, indexes Serena only if the verdict says yes, generates layer specialists via /adapt-to-stack, verifies the tracker adapter, writes two memories, then runs every check. - Strict write boundary: works autonomously inside the repo but stops and asks before any write outside it, including memory writes and allowlist edits. - Idempotent re-runs: every step checks the disk and skips itself, so a crashed session resumes by simply running /bootstrap again with no progress files. - Use Case: After charting closes with a named stack and checks, run /bootstrap to make the repo real and get a seven-row evidence-backed report classifying each seam check as pass or fail. ## Quick Start Run /bootstrap on the empty repo that charting just finished deciding, and wait for the full pass/fail report before committing anything.

Frequently Asked Questions about bootstrap

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

FAQPage Schema
How do I bootstrap an empty repo after charting?

Run /bootstrap once charting has closed with a named stack, layer chain, Serena verdict, and bootstrap checks. The skill scaffolds the stub, writes CLAUDE.md, generates specialists, and runs every check, ending with one pass/fail report.

What must exist before running the bootstrap step?

Five inputs are required: an existing empty repo, a named stack, a named layer chain, the Serena verdict, and the bootstrap checks from charting's tail. If any is missing, the skill stops and says charting is not finished rather than choosing for you.

Does bootstrap decide the tech stack for me?

No. Bootstrap decides nothing; every choice comes from the prior charting stage. If something genuinely needs deciding, it stops and flags it as a human decision rather than filling the gap.

Can I re-run bootstrap after a session crashes?

Yes. Every step checks the disk and skips work already done, so running /bootstrap again resumes safely. No progress file is written because the disk itself is the source of truth.

When does bootstrap index Serena for a repo?

Only when the Serena verdict recorded in CLAUDE.md says yes. If the verdict is no, the step is skipped and recorded as 'skipped — verdict no', which counts as a pass, not a gap.

What happens when a bootstrap check fails?

The skill runs all checks anyway, reports every red with evidence, commits nothing, and stops. Classifying a failure as a wrong check or a wrong stack is left to the human, since that decision may reopen the map.