Claude Skills + Happy Hour (Zoom)

For our first event, let's meetup on Zoom. Bring your own drink/beer as our first community event.

Update: Here's the Skill Library that we'll be working with tonight:
https://github.com/Claude-AI-Tools-Ventura-County/giant-brains-claude-skills

Most people use Claude like a search engine — one question, one answer, done. Skills change that. A Claude Skill is a reusable, trigger-aware prompt that turns Claude into a specialist: a cost auditor, a decision framer, a plan stress-tester, a bottom-line extractor.

In this session we'll show you exactly what a Skill looks like under the hood, demo three live Skills that will immediately change how you work, and build one from scratch together — in the room, in real time. You'll leave with a working Skill you can drop into Claude Code, claude.ai, or any AI workflow the same day. No prior Claude experience required. Bring a problem you wish Claude handled better.

• Hook — The problem (3 min). Show Claude giving a bloated answer. Ask: "what if Claude had a mode for this?" Introduce the skill concept in one sentence.
• What a skill is (5 min). Show a raw SKILL.md (e.g. `bottom-line`). Anatomy: trigger description, rules, output format. Skills live in files, not in your head.
• Live demo — 3 skills in sequence (15 min).
• `bottom-line` → paste a long AI response, invoke skill, see the trim.
• `blast-radius` → propose a schema migration, watch the cost surface.
• `iron-triangle` → "we need to ship by Friday," watch the tradeoff framing.
• How triggering works (5 min). Skill descriptions tell Claude when to fire. Show auto-trigger vs. explicit invocation. Brief mention of CLAUDE.md / AGENTS.md as the registry.
• Write your first skill live (8 min). Pick a pain the audience names. Draft a SKILL.md together: name, description trigger, 3-5 rules, output format. Ship it.
• Q&A + takeaway (4 min). Point to Giant Brains repo. Challenge: one skill written and tested before next meetup.