What problem does it solve?
Greenfield repo planning can be time-consuming and error-prone; this skill interviews the user to determine intended AI CLIs (Claude Code, OpenCode, Kilo Code, Codex, Gemini), languages, test frameworks, deployment targets, and compliance/workflow requirements for a new repo, then delegates to skill-sync to install the matching subset into each CLI's per-repo directory. Use when starting a fresh repo that has no code yet — for existing repos, use skill-sync directly (its detection is more accurate than any interview).
Core Features & Use Cases
- AskUserQuestion-driven interview to define repo goals and constraints across CLIs and tools.
- Produce a per-repo .skill-sync.yaml allowlist that guides per-repo installations via skill-sync.
- Not intended for existing repos; if code markers exist, run skill-sync directly for deterministic provisioning.
- Delegate provisioning to skill-sync to install the selected CLI subsets into per-repo locations.
Quick Start
Run this skill on a new repository to interview the user, generate a per-repo skill allowlist via skill-sync, and install the matching subset into each CLI's per-repo directory.