Warn on NaN passed to aria-* attributes — runtime detection & fixes

NaN values passed to aria-* attributes produce subtle accessibility failures—screen readers can ignore or misannounce controls—but these issues are hard to find with static analysis or post-release audits because the NaNs frequently arise only at runtime. The burden falls on front-end teams building dynamic UIs—enterprise React shops, accessibility engineers, QA teams, and compliance/legal teams—where a small number of bad attributes can create outsized remediation costs and regulatory exposure. A practical product would provide a lightweight runtime detector that warns (and can optionally sanitize) NaN values in development and CI, paired with an ESLint/Babel plugin and code-action remediation suggestions, plus a small incidents dashboard for prioritization. Shipping the core as an open-source, React-first library with an opt-in commercial CI plugin and remediation analytics lowers adoption friction while enabling monetization. This market is attractive now: an addressable base of roughly 10M web development teams spending about $2,400 per year on tooling and accessibility implies a $24B opportunity, and increasing regulatory pressure and a shift-left mentality mean teams are buying dev-time and CI-first checks. Framework consolidation around React further accelerates reach for a focused offering, while companies are explicitly investing in measurable a11y risk reduction. Differentiation comes from highly accurate, low-overhead runtime checks that prioritize actionable fixes over noisy alerts and from deep IDE/CI integrations that can demonstrate measurable risk reduction to compliance stakeholders. Given the market size and clear go-to-market path, this is worth pursuing if the team can deliver a performant, low-noise React-first runtime and secure early enterprise adopters; the main risks are minimizing false positives, keeping production impact negligible, and managing complexity when expanding beyond React.