Applications
An on-demand crypto market intelligence and decision-support terminal powered by TypeSafe AI's System One model (Jev).
from typesafe_sdk import AsyncTypeSafeClient, Choice, Noul, Scoreapp/services/typesafe_service.py:3
The link points at the commit we read, so the line number still holds.
These question sets are lifted from this project's source exactly as written.
from app/services/typesafe_service.py:49
sentiment_spectrumscoreRate the prevailing social mood in `social_stats` and `representative_tweets`.
0 — Extreme Panic / Capitulation1 — Cautious / Bearish2 — Neutral / Mixed3 — Optimistic / Bullish4 — Euphoric / Greedycatalyst_impactscoreRate the significance of any events or breaking news described in `representative_tweets`.
0 — No news or pure retail noise1 — Minor routine update or rumors2 — Moderate ecosystem milestone3 — Major market-shifting catalystimport { experimental_evaluate as evaluate } from 'ai';
const { answers } = await evaluate({
model: 'typesafe-ai/jev',
state,
questions: {
sentiment_spectrum: { type: 'score', instructions: 'Rate the prevailing social mood in `social_stats` and `representative_tweets`.',
criteria: ['Extreme Panic / Capitulation', 'Cautious / Bearish', 'Neutral / Mixed', 'Optimistic / Bullish', 'Euphoric / Greedy'] },
catalyst_impact: { type: 'score', instructions: 'Rate the significance of any events or breaking news described in `representative_tweets`.',
criteria: ['No news or pure retail noise', 'Minor routine update or rumors', 'Moderate ecosystem milestone', 'Major market-shifting catalyst'] },
},
});