I maintain Adanos, a commercial market-sentiment API with a free tier. Would Folio welcome one optional, read-only US-equity social-sentiment capability?
I inspected the capability registry, canonical instrument work (#23), and the existing market.sentiment manifest. That existing capability is Longbridge market temperature for US/HK/CN/SG; I would leave it unchanged. A separate research.assetSentiment (or your preferred name) could accept a resolved US stock instrument, explicit UTC window, and source selector (Reddit, X/FinTwit, News, Polymarket), return structured metrics with provider/time provenance, and preserve missing or plan-restricted results rather than inventing neutral scores. No trading logic or portfolio access.
The narrow first PR would add the provider-backed capability, mocked tests and setup docs. Before coding, could you confirm: (1) the preferred capability ID/category, (2) where a user-supplied Adanos key should live in the current local credential model, and (3) whether the initial capability should be agent-visible, UI-visible, or both? I will not repurpose the existing market-temperature API.
I read CONTRIBUTING.md and will keep the eventual PR small, with reproducible Bun/test results and screenshots if any visible UI changes are included.
I maintain Adanos, a commercial market-sentiment API with a free tier. Would Folio welcome one optional, read-only US-equity social-sentiment capability?
I inspected the capability registry, canonical instrument work (#23), and the existing
market.sentimentmanifest. That existing capability is Longbridge market temperature for US/HK/CN/SG; I would leave it unchanged. A separateresearch.assetSentiment(or your preferred name) could accept a resolved US stock instrument, explicit UTC window, and source selector (Reddit, X/FinTwit, News, Polymarket), return structured metrics with provider/time provenance, and preserve missing or plan-restricted results rather than inventing neutral scores. No trading logic or portfolio access.The narrow first PR would add the provider-backed capability, mocked tests and setup docs. Before coding, could you confirm: (1) the preferred capability ID/category, (2) where a user-supplied Adanos key should live in the current local credential model, and (3) whether the initial capability should be agent-visible, UI-visible, or both? I will not repurpose the existing market-temperature API.
I read CONTRIBUTING.md and will keep the eventual PR small, with reproducible Bun/test results and screenshots if any visible UI changes are included.