← ClaudeAtlas

brief-melisted

Run Shiva's pre-project vocabulary briefing for any GenAI build or technical workflow. Use when the user invokes /shva:brief-me, says "brief me", asks for a SHVA brief, or asks to understand a domain before building. Produces a 20-25 term glossary, 3 sharpening questions, correction coda, and a mandatory HTML visual mental model unless the user explicitly opts out.
Shivak11/shva-marketplace · ★ 0 · AI & Automation · score 60
Install: claude install-skill Shivak11/shva-marketplace
# SHVA Brief-Me Skill Use this skill when the user asks for `/shva:brief-me`, `brief-me`, a SHVA brief, or a pre-project vocabulary briefing. ## Required Output Deliver all of these, in this order: 1. A 20-25 term glossary in three buckets: - Creative / Style: aesthetic vocabulary, mood, tone, composition, genre. - Technical / Process: engines, formats, controls, mechanics, lifecycle. - Prompt / Direction: terms the user can actually use in future prompts. 2. Exactly 3 clarifying questions: - one about aesthetic/reference points, - one about scope/fidelity, - one about constraints. 3. The correction coda: > "Now, given these terms, here's what I'm building: **[your idea]**. > Flag any terms I'm using incorrectly and suggest better vocabulary > before you write any code or output." 4. A visual mental model as an HTML explainer file, unless the user explicitly says `no visual`, `glossary only`, `skip the diagram`, or similar. ## HTML Visual Hard Gate The visual is not optional by default. Do not finish a `brief-me` request with only prose, a widget call, or a chat summary. Before final response, verify that one of these is true: - an HTML explainer file was created and linked in the final response, or - the user explicitly opted out of the visual. If the current project is writable, create the file under `docs/` using a descriptive slug, for example: `docs/<topic>-brief.html` If the current project is not writable, create it under `/private/tmp/`