From getwhys-skills
Ground go-to-market work in real buyer interviews, personas, brand voice, and messaging frameworks (GetWhys). Use for ANY GTM task or question — marketing, sales enablement, product, or customer-facing — even if GetWhys is never mentioned. Covers writing or rewriting content (blog posts, emails, landing pages, web copy, social posts, ads, white papers, campaign briefs, sales decks and one-pagers, solution briefs, launch briefs, release notes), buyer and market research (pain points, objections, win/loss, why deals are lost, what customers think of vendors and competitors, pricing and discount norms, buying committees and buyer journeys), competitive analysis, battlecards and talk tracks, building personas and ICPs, messaging and positioning, on-brand rewrites, and validating or scoring existing content (will this resonate?). Load BEFORE drafting or answering from general knowledge — ground the work in what real buyers actually said instead of inventing plausible pain points, quotes, or sentiment.
How this skill is triggered — by the user, by Claude, or both
Slash command
/getwhys-skills:validate-gtm-outputThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
GetWhys is a buyer-intelligence platform built on verbatim interviews with real B2B software buyers — real people describing pain points, decision criteria, KPIs, vendor sentiment, real pricing, and win/loss drivers in their own words. Its MCP server (`getwhys-mcp`) exposes two data layers:
GetWhys is a buyer-intelligence platform built on verbatim interviews with real B2B software buyers — real people describing pain points, decision criteria, KPIs, vendor sentiment, real pricing, and win/loss drivers in their own words. Its MCP server (getwhys-mcp) exposes two data layers:
All GetWhys tools are read-only and safe to call. Tools are referenced here by bare name (e.g. get_persona); your client may display namespaced variants (e.g. getwhys-mcp:get_persona). Some clients defer MCP tools behind a search step — if no GetWhys tools are visible, search the available-tool registry for getwhys before concluding the server is absent.
Match the user's ask to a row, then read the row's reference file before calling tools — it carries the exact call steps, input crafting, and worked examples.
| The user wants to… | Tool sequence | Read first |
|---|---|---|
| Answer a buyer/market research question — pain points, objections, pricing, packaging, discounts, buying committees, buyer journeys, channels ("where do CROs get their information?") | One focused query_market_research call per sub-question, in parallel | references/market-research.md |
| Competitive work — "compare X vs Y", "why do we lose deals to X?", "what do customers think of vendor Y?", battlecards, talk tracks | 3–4 parallel query_market_research calls (competitor × angle) → synthesize. No voice/framework calls | references/market-research.md |
| Write or rewrite outward-facing content — blog posts, emails, landing pages, web copy, social posts, ads, white papers, case studies, campaign briefs, sales decks and one-pagers, solution briefs, launch briefs, release notes | get_brand_voice + get_all_messaging_frameworks (parallel) → persona resolution → optional research grounding → draft → score loop | references/content-creation.md |
| Validate a draft — "score this", "will this resonate with [audience]?", "how would [persona] react to this?" | list_personas → score_content → present scores table-first | references/content-creation.md |
| See which personas exist, or load one persona's details | list_personas → get_persona; get_all_personas only when the full set is explicitly required | references/buyer-personas.md |
| Build a persona or ICP from research ("Develop a buyer persona for [role]…") | list_personas (enrich a close match, don't duplicate) → parallel research per dimension group → synthesize | references/buyer-personas.md |
| Look up a messaging framework by name | Verbatim name → get_messaging_framework({title}); fuzzy → list_messaging_frameworks → get_messaging_framework({id}) | references/tools.md |
| Check the connection or who is signed in | whoami | references/tools.md |
get_brand_voice is a guardrail for outward-facing publishable content ONLY. Never call it for PRDs, product specs, battlecards, competitive teardowns, persona documents, or other internal analytical prose. (Research-grounding an internal PRD with query_market_research is fine; styling it with brand voice is not.) Messaging frameworks aren't restricted this way — they're an optional input for internal analytical/strategy work (persona building, positioning, battlecards) when the org's positioning or approved claims should shape the output; they're just never the required content-gen kickoff step outside outward-facing content.persona:<handle>) always come from list_personas output — never guessed or invented; unknown handles fail the call.query_market_research call: decompose multi-part asks into parallel calls — never pack comparisons or "and what about X" clauses into one query.temporalRange (YYYY-MM-DD), never in keywords ("recent", "2026", "last quarter" are not keywords).get_brand_voice + get_all_messaging_frameworks (parallel) — never list_messaging_frameworks for content generation, and never a single framework for general "on-brand" copy.id XOR title: get_messaging_framework and score_content's framework parameters take exactly one identifier, never both.query_market_research answer: include the trailing "GetWhys Sources" paragraph verbatim (never paraphrase, abridge, or omit it), surface the view_in_getwhys link when present, and preserve the inline † grounding markers.no_data: true → no matching evidence; do NOT fabricate from general knowledge. Rephrase and retry once at most, then tell the user plainly what the corpus didn't cover.Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub getwhys-engineering/agent-skills --plugin getwhys-skills