agwacom/avad-help

<!-- AUTO-GENERATED from SKILL.md.tmpl — do not edit directly --> <!-- Regenerate: bun run gen:skill-docs -->

/avad-help — Skill Hub & Update Check

When invoked, show all available avadbot skills, help the user pick the right one, and check for updates.

Step 1: Version Check

Check for updates by comparing the installed version against the latest release:

# Get installed version from plugin.json
PLUGIN_DIR=$(find ~/.claude/plugins/cache -path "*/avadbot*/.claude-plugin/plugin.json" 2>/dev/null | head -1)
INSTALLED=$(cat "$PLUGIN_DIR" 2>/dev/null | grep '"version"' | head -1 | sed 's/.*"version": *"\([^"]*\)".*/\1/' || echo "unknown")

# Check latest version from GitHub
LATEST=$(gh api repos/agwacom/avadbot/releases/latest --jq '.tag_name' 2>/dev/null | sed 's/^v//' || echo "")
# Fallback: check package.json on main branch
if [ -z "$LATEST" ]; then
  LATEST=$(gh api repos/agwacom/avadbot/contents/package.json --jq '.content' 2>/dev/null | base64 -d | grep '"version"' | head -1 | sed 's/.*"version": *"\([^"]*\)".*/\1/' || echo "unknown")
fi

If update available: Show a notice at the top:

** Update available: v{INSTALLED} → v{LATEST}
   Run: claude plugin update avadbot

If up to date or check fails: Skip silently.

Step 2: Show Skill Catalog

Display all available skills in a table:

avadbot v{VERSION} — Skill Catalog

Skill                        What it does                              When to use
─────────────────────────── ──────────────────────────────────────── ──────────────────────────────────
/avadbot:avad-plan-ceo-review  CEO/founder plan review — rethink the    Before building: challenge the
                               problem, find the 10-star product        problem, expand/reduce scope

/avadbot:avad-plan-eng-review  Eng manager plan review — architecture,  Before building: lock in the
                               data flow, edge cases, test coverage     execution plan

/avadbot:avad-review           Pre-landing code review with Fix-First   Before shipping: review your
                               — AUTO-FIX mechanical issues             diff, fix issues

/avadbot:avad-ship             One-command ship — validate, sync,       When ready to ship: push +
                               test, review, push, create PR            create PR automatically

/avadbot:avad-qa               QA test a web app — find and fix bugs,   After deploying: systematic
                               generate regression tests                testing + bug fixing

/avadbot:avad-debug            Systematic root-cause debugger —         When stuck on a bug: traces
                               Iron Law, no fix without root cause      data flow, tests hypotheses

/avadbot:avad-document-release Post-ship doc updater — syncs README,    After shipping: update all
                               CHANGELOG, CLAUDE.md with what shipped   docs to match reality

/avadbot:avad-retro            Weekly engineering retrospective —       End of week: commit analysis,
                               commit analysis, team breakdown          patterns, improvements

/avadbot:avad-browse            Headless browser — navigate, click,      When you need to see a page:
                               screenshot, test forms, import cookies   QA, dogfooding, verification

/avadbot:avad-plan-design-review Designer's eye plan review — rates 8     Before building UI: design
                               design dimensions, produces fix plan     critique and score

/avadbot:avad-design-consultation Design system creation — competitive   Starting a project: create
                               research, DESIGN.md, color/font system   DESIGN.md with design system

/avadbot:avad-design-review    Visual QA + fix loop — spacing, slop,   After building UI: find and
                               interaction issues, risk calculator      fix visual issues

/avadbot:avad-careful          Destructive command guardrails — warns   Safety: prevent accidental
                               before rm -rf, DROP TABLE, force-push   destructive commands

/avadbot:avad-freeze           Restrict edits to a directory — locks   Focus: prevent changes outside
                               Edit/Write to a specific path           the directory you're working in

/avadbot:avad-unfreeze         Clear freeze boundary                   After /avad-freeze: unlock edits

/avadbot:avad-guard            Combined careful + freeze in one cmd    Safety + focus: both guardrails

/avadbot:avad-new              Project scaffolding — asks 5 product      Starting a project: scaffold
                               questions, generates 16 pre-filled       16 pre-filled files
                               files, detects stack, creates repo

/avadbot:avad-help              Skill hub — shows all skills,            When unsure which skill to use:
                               context-aware routing, update check      see what's available

/avadbot:avad-hello            Test greeting — confirms plugin works    Testing: verify installation

Step 3: Context-Aware Recommendation

If the user described what they want to do (either in the invocation or as a follow-up), recommend the best skill:

Role-based routing:

• "CEO", "founder", "product" → /avadbot:avad-plan-ceo-review
• "eng", "engineer", "architect", "technical" → /avadbot:avad-plan-eng-review
• "review", "diff", "review PR", "code review" → /avadbot:avad-review
• "ship", "push", "deploy", "merge", "create PR" → /avadbot:avad-ship
• "QA", "test", "bug", "broken" → /avadbot:avad-qa
• "debug", "fix", "stuck", "error", "crash" → /avadbot:avad-debug
• "docs", "documentation", "README", "changelog" → /avadbot:avad-document-release
• "retro", "retrospective", "week", "metrics" → /avadbot:avad-retro
• "browse", "browser", "screenshot", "page" → /avadbot:avad-browse
• "design plan", "design critique", "design score" → /avadbot:avad-plan-design-review
• "design system", "DESIGN.md", "colors", "typography" → /avadbot:avad-design-consultation
• "visual QA", "design audit", "spacing", "AI slop" → /avadbot:avad-design-review
• "new project", "scaffold", "init", "start a project" → /avadbot:avad-new
• "careful", "safety", "prod mode", "guardrails" → /avadbot:avad-careful
• "freeze", "restrict edits", "lock directory" → /avadbot:avad-freeze
• "unfreeze", "unlock edits" → /avadbot:avad-unfreeze
• "guard", "full safety" → /avadbot:avad-guard

If no context given: Use AskUserQuestion: "What are you trying to do?" with options based on common workflows:

• A) Review a plan before building
• B) Review code before shipping
• C) Ship my changes (push + PR)
• D) Test/QA a web app
• E) Debug a bug or error
• F) Update docs after shipping
• G) Browse/screenshot a page
• H) Run a weekly retro
• I) Design system or design review
• J) Start a new project (scaffold files)
• K) Enable safety guardrails (careful/freeze/guard)

Then recommend the specific skill with the exact command to run.

Step 4: Workflow Suggestions

After recommending, briefly show the typical workflow order:

Typical workflow:
  New  →  Plan  →  Build  →  Review  →  Ship  →  QA  →  Docs  →  Retro
  /avad-new  →  /avad-plan-*  →  /avad-review  →  /avad-ship  →  /avad-qa  →  /avad-document-release  →  /avad-retro

Rules

• Keep output concise — one screen max for the catalog
• Never run any other skill — only recommend which to use
• If the user already knows which skill they want, just confirm and show the command
• Update check failure is silent — never error on network issues

Agent Identity

Use your full model name and version as your agent identifier in all GitHub output. Examples: "claude Opus 4.6", "codex o3", "gemini 2.5 Pro"

GitHub Output

When creating or commenting on GitHub issues, discussions, or pull requests:

1. Title prefix: Include the skill role in the title. Format: [avad-help] {short description} Example: [avad-help] skill catalog and update check

2. Labels:

   • Role label: avad-help
   • Add any project-specific labels per the project's labeling convention.
   • Do NOT add task/work labels — help output is informational, not tasks.
   • Create the label if it does not exist (color: #0E8A16, description: "Skill hub and update check").

3. Both title prefix and label are required.

Headings

Prefix all comment, issue, discussion, and PR section headings with your agent identity. Example: "## claude Opus 4.6 Skill Catalog"

Signature

End every GitHub issue, discussion post, PR description, review comment, or review response with a signature line:

---
_Generated by {agent identity} · /avad-help · {YYYY-MM-DD}_

Example: _Generated by claude Opus 4.6 · /avad-help · 2026-03-19_