This repo now serves the Qwantify Hyperliquid analytics workspace on the existing Cloudflare Pages project.
Deploys continue targeting the existing hl-stats Pages project, so the production URL stays the same.
- Local app dev:
npm run dev - Pages build output:
npm run pages:build - Pages local preview:
npm run pages:dev - Pages deploy:
npm run pages:deploy
/app/research absorbs the deterministic market-factor research tool from the
AI Hedge Fund project into the Hyperliquid interface. It keeps the documented
equities/SPY methodology, normalized price chart, factor attribution, evidence
dates, headline metadata, and browser-local run history. It does not place
trades or expose provider credentials to the browser.
The provider adapter is a same-origin Cloudflare Pages Function written for the Workers runtime. For local development:
cp .dev.vars.example .dev.vars
# Add the local EODHD token to .dev.vars, then:
npm run devFor the deployed hl-stats Pages project, configure the token as an encrypted
secret before deployment:
npx wrangler pages secret put EODHD_API_TOKEN --project-name=hl-stats
npm run pages:deploySame-origin request checks are not user authentication. Treat the deployed research endpoint as public, and put the Pages project behind Cloudflare Access or add an abuse-protection policy before relying on it for a paid EODHD allowance.
The interactive research path is deliberately bounded to one ticker, one SPY benchmark series, and optional recent headlines per request. Larger multi-year, multi-ticker backtests belong in a separate asynchronous Workflow rather than a Pages request.