gylt-new-addon

name

gylt-new-addon

description

Scaffold a new GYLT addon — a standalone repo of skills, hooks and extras that plugs onto GYLT. Generates the folder structure, the manifest (addon.json), a ready-to-run installer, a README and a LICENSE from templates, plus one starter skill. Invoke when you want to package and share a set of Claude Code skills/hooks as an installable addon (e.g. a dev toolkit, a writing toolkit) on top of GYLT. GYLT is the base; addons extend it.

/gylt-new-addon

GYLT is the base. Addons extend it: a standalone git repo holding skills, hooks and extra files, installed with one script. This skill scaffolds a new addon from templates so every addon shares the same shape and installer — no copy-paste from an existing one.

Reference addon (the shape this produces): GYLT-DEV-CONTEXT-ADDON.

Step 1 — Gather the addon spec

Ask the user (only for what is missing from the conversation):

• Name — kebab-case, e.g. gylt-writing-addon. Becomes the repo/folder name and addon.json.name.
• Description — one line, what the addon gives.
• Target directory — where to create it. Default: ~/Dev/<name>.
• First skill name — kebab-case; a starter skill is scaffolded so the addon is non-empty.

Confirm the resolved plan (name + path) before writing anything.

Step 2 — Create the structure

<target>/
  addon.json
  install.sh
  README.md
  LICENSE
  skills/
    <first-skill>/SKILL.md
  hooks/                  (empty — add hook scripts here)
  project-context-tools/  (only if the addon extends prepare-project-for-agent; else omit)

mkdir -p "<target>/skills/<first-skill>" "<target>/hooks"

Step 3 — Fill from templates

The templates live next to this skill, in templates/:

Template	→ Output	Substitutions
templates/addon.json	<target>/addon.json	{{NAME}}, {{DESCRIPTION}}, {{SKILLS}} (JSON array), {{HOOKS}} (JSON array, [] if none)
templates/install.sh	<target>/install.sh	none — it is generic, it reads addon.json at runtime
templates/README.md	<target>/README.md	{{NAME}}, {{DESCRIPTION}}
templates/SKILL.md	<target>/skills/<first-skill>/SKILL.md	{{SKILL_NAME}}, {{SKILL_DESCRIPTION}}

For the LICENSE, copy GYLT's LICENSE (MIT) verbatim, keeping the copyright holder.

After substitution, strip every {{...}} placeholder — none must remain in the output.

addon.json.skills lists the skill folder names; addon.json.hooks lists { "file": "hooks/<x>.sh", "event": "<SessionStart|...>" } objects (empty array if the addon ships no hooks).

Step 4 — Make scripts executable & verify

chmod +x "<target>/install.sh"
bash -n "<target>/install.sh"      # syntax check

Confirm the tree to the user and show the install command:

cd <target> && bash install.sh

Step 5 — Next steps for the user

Tell them:

• Add real skills under skills/<name>/SKILL.md (folder-per-skill; templates can sit beside SKILL.md).
• Add hook scripts under hooks/ and declare them in addon.json.hooks.
• git init, push, and share — others install with bash install.sh.

Rules

• One addon = one repo. Skills are folder-per-skill (skills/<name>/SKILL.md) so they can carry sibling templates.
• The installer is generic and reads addon.json — never hand-edit the install logic per addon; if it needs to change, change the template here so all future addons benefit.
• Do not bundle anything proprietary or non-shareable into a public addon.