Mission: Forge the Domain Expansion (Hard • 8 pts)
Design a new summarizer agent that fuses retrieval-ranked passages, market movement signals, and bias flags into a structured, cited summary.
Goals
- Build
hybrid_rag_summarizer agent:
- Ingest:
market_data[], news_articles[]
- Retrieve and rank passages (reuse ranking from Issue 5 if available)
- Compute price deltas per symbol and classify move regimes (flat/up/down)
- Integrate bias flags as uncertainty sources
- Output structured JSON:
{ summary, key_points[], uncertainty_factors[], citations[] }
Where
- New agent file:
Jujutsu-Quants/app/adk/agents/hybrid_rag_summarizer.py
- Wire into orchestrator optional stage (behind a flag in
ADK_CONFIG)
Minimal Reproducible Example (MRE)
uvicorn Jujutsu-Quants.app.adk.main:app --reload
curl -s -X POST http://localhost:8000/api/v2/report -H "Content-Type: application/json" -d '{
"symbols": ["AAPL"],
"news_urls": ["https://example.com/news1"],
"question": "What are key drivers today?"
}' | jqExpected top-level include key rag_summary:
{
"rag_summary": {
"summary": "Apple rose modestly; supply chain easing ...",
"key_points": ["iPhone demand", "China recovery"],
"uncertainty_factors": ["single-source report", "political risk"],
"citations": [{"source": "https://example.com/news1", "start": 120, "end": 220}]
}
}Acceptance Criteria
- New agent produces deterministic structure and integrates three signals (price regime, ranked passages, bias flags)
- Orchestrator can include/exclude via config
- Unit tests for passage fusion and regime classification
- Docs updated (README agents section) with a short usage note
Mission: Forge the Domain Expansion (Hard • 8 pts)
Design a new summarizer agent that fuses retrieval-ranked passages, market movement signals, and bias flags into a structured, cited summary.
Goals
hybrid_rag_summarizeragent:market_data[],news_articles[]{ summary, key_points[], uncertainty_factors[], citations[] }Where
Jujutsu-Quants/app/adk/agents/hybrid_rag_summarizer.pyADK_CONFIG)Minimal Reproducible Example (MRE)
Expected top-level include key
rag_summary:{ "rag_summary": { "summary": "Apple rose modestly; supply chain easing ...", "key_points": ["iPhone demand", "China recovery"], "uncertainty_factors": ["single-source report", "political risk"], "citations": [{"source": "https://example.com/news1", "start": 120, "end": 220}] } }Acceptance Criteria