Builds new, user-owned capabilities for the ALIVE Context System: repeatable skills, behavioral rules, and automated hooks. It guides the agent through compatibility checks, writing files into the human's workspace (not plugin cache), creating discovery symlinks, and validating the new capability so it loads reliably.
Use Extend when you want the agent to create a persistent, sharable capability (a SKILL.md, rule, or hook) rather than a transient preference change. It's appropriate for automations that should survive plugin updates, for packaging capabilities for contributors, and when the agent must verify compatibility with plugin and hook APIs.
Likely used by Claude Code / Claude plugin host environments and agents that can write files and create symlinks (Claude Code, local CLI-based agents).
Documentation-only skill for the ALIVE Context System that guides agents in creating custom skills, rules, and hooks. No scripts to run — purely instructional. Well-structured SKILL.md with clear flow, concrete file paths, and good separation of extension types. Scoped to ALIVE ecosystem users.
Clean skill, no security concerns. No scripts, no shell execution, no network calls. Documentation quality is good but could include error handling guidance. Usefulness is limited by ecosystem specificity.