HuskyDanny Sentiment Analysis Workflow
OBJECTIVE: Determine overall news sentiment and key themes for the symbol.
Step 1: Gather Recent News
- Use
get_news_sentimentwith max_results=10 - Focus on news from the past 7 days
Step 2: Classify Each Item
For each news item, classify as:
- POSITIVE: Bullish news (earnings beat, upgrades, product launches)
- NEGATIVE: Bearish news (misses, downgrades, lawsuits, delays)
- NEUTRAL: Informational without clear sentiment
Step 3: Calculate Aggregate Score
- Count: X positive, Y negative, Z neutral
- Score = (positive - negative) / total
- Score > 0.3: BULLISH sentiment
- Score < -0.3: BEARISH sentiment
- Otherwise: MIXED sentiment
Step 4: Identify Themes
- What topics appear repeatedly?
- Any developing narratives?
- Institutional vs retail focus?
Output Format
Sentiment Score: [+X.XX or -X.XX] -> [BULLISH/BEARISH/MIXED] Distribution: {positive} positive, {negative} negative, {neutral} neutral Key Themes:
- [Theme 1 with example headline]
- [Theme 2 with example headline] Dominant Narrative: [Summary of overall story]