install-apm-package Skill 🚀

Overview

This skill manages the safe deployment of APM primitives into their respective runtime environments. It enforces validation and target discovery to ensure byte-identical reproducibility across environments.

🚫 Non-Negotiables

1. Source Integrity — Never edit files in .agents/, .github/, or .claude/ directly. Always edit the source in .apm/ and re-install.
2. Lockfile Persistence — apm.lock.yaml MUST be committed after any install change.
3. Validation First — Always run validate_apm_package.py before executing the install.
4. Project Root Discipline — Run apm install from the project root to ensure converged skills land in the authoritative .agents/ folder. Running inside a package directory creates local, isolated artifacts.

Decision Tree

1. Global Activation? -> Run apm install ./path-to-package from the project root.
2. Isolated Testing? -> Run apm install from within the package directory.
3. Target Inferred? -> (e.g., .claude/ exists) -> Run apm install.
4. Ambiguous Target? -> (No signals) -> apm install will exit with code 2; ask user for explicit --target (e.g., agent-skills, claude, all).
5. Converged Skills? -> (Standard) -> Deploys to .agents/skills/. Use --target agent-skills for explicit skill deployment.
6. Legacy Skills? -> (Per-client paths) -> Use --legacy-skill-paths flag.
7. Production/CI? -> Use --frozen flag to ensure lockfile compliance.