xuenerissa-png
UserDisciplined competitor research SOP. A Claude Skill that refuses lazy AI shortcuts — every funding number needs 2 sources, every customer overlap needs verification, every recommendation needs 6-dimension scoring.
Categories
Indexed Skills (3)
nerissa-competitive-research
A disciplined 7-step competitive research SOP that prevents lazy AI shortcuts (sourcing from secondhand blogs, inheriting unverified judgments, confusing funding size with relevance). Forces verifiable two-source funding data, official-site visits, essence decomposition (technology vs packaging-layer vs sales), 6-dimension scoring, and dual-track separation (fundraising reference vs business learning target). Use when user says "research X company", "do competitive analysis", "调研 X 家", "竞品研究", "stress-test this competitor", or when adding any new competitor to a comparison table.
nerissa-self-audit
A disciplined 9-lens framework for auditing your own startup's current state — vision vs reality, customer fit, business model coherence, resource allocation, hidden assumptions. Forces you to apply First Principles + 8 classic strategy frameworks (JTBD / 5 Whys / OKR / SWOT / Porter / BCG / Blue Ocean ERRC / Inversion) to your own company, not just competitors. Use when user says "audit our current state", "我们现在到底是什么", "诊断我们自己", "review our positioning", "self-audit", "where are we really", or when vision and reality seem misaligned.
decision-lens
Pressure-test complex decisions through multiple independent expert lenses. Forces disagreement to surface and converges on a recommendation plus the questions only you can answer. Use when the user says "review this", "let's discuss", "help me think through", "multi-role review", "评审", "审视", "多角色过一下", or when facing complex decisions about product direction, architecture, scope, positioning, or cross-project tradeoffs.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.