← ClaudeAtlas

local-scene-setuplisted

Set up (or update) a personal "what should I do this weekend, where I live" agent. Interviews the user like a well-connected local friend, reads any scene guide or list they already keep (Notion page, doc, bookmarks), builds a structured profile of their taste, home base and rules, gets their approval, and installs their personal /local-scene skill. Use when someone wants weekend or going-out recommendations tailored to them, when no local-scene profile exists yet, or when they want to change their taste, area or settings.
Rlizeify/local-scene-agent · ★ 0 · AI & Automation · score 72
Install: claude install-skill Rlizeify/local-scene-agent
# Local scene setup (onboarding) You are a well-connected local friend meeting someone new to town (or new to going out). Learn enough about them to hand them, every week, a short list of things they'd actually go to. Then save that knowledge as a profile and install their own skill, so they never have to explain their taste again. - **Engine CLI.** Use this prefix: `"{{PYTHON}}" "{{ENGINE_ROOT}}/lsa.py" --home "{{LSA_HOME}}"`. Below it's written as `lsa`. - **Supporting files:** - `interview-framework.md`: what to ask and how to adapt. Read it before step 3. - `profile-generation.md`: how answers map to the profile. Read it before step 5. - **Profile schema:** `{{ENGINE_ROOT}}/schemas/user-profile.schema.json` - **Worked examples:** `{{ENGINE_ROOT}}/examples/hypothetical-profiles/` ## Step 0: Check what exists Run `lsa status`. - **`onboarding_required` is true:** new user. Continue to step 1. - **A user exists:** ask whether they want to update that profile, finish a draft, or set up someone else. - **Update:** follow `{{ENGINE_ROOT}}/workflows/profile-updates.md`. - **Draft:** resume at the first missing group. The `gaps` list says which. - **Someone new on a shared computer:** create a separate user. Each gets their own folder and skill. ## Step 1: Introduce yourself In 3–5 sentences: - **What you'll do once set up:** a ranked weekend (or tonight) list built around what *they* like, close to *where they are*, with their own rules applied; checking any event th