charieswei
/
TradingAgents
mirror of https://github.com/TauricResearch/TradingAgents.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Add Polaris as news/sentiment/price data vendor 

Adds Polaris Knowledge API as a third data vendor alongside yfinance and
Alpha Vantage. Polaris provides sentiment-scored intelligence, composite
trading signals, and news impact analysis that raw data feeds don't offer.

Implements all existing vendor methods plus 3 Polaris-exclusive methods
(sentiment_score, sector_analysis, news_impact).

Uses cachetools.TTLCache for thread-safe caching, shared financials fetch
to avoid redundant API calls, and polaris-news SDK for all API access.

Tested with real API calls — all functions return valid data.
Addresses #305, #11, #86.
This commit is contained in:
John Weston 2026-03-23 08:51:57 -04:00
parent 589b351f2a
commit 4abad33e0f
4 changed files with 632 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
requirements.txt
View File

@ -1 +1,3 @@
.
polaris-news>=0.6.0
cachetools>=5.0.0

42
tradingagents/dataflows/interface.py
View File

@ -23,6 +23,20 @@ from .alpha_vantage import (
get_global_news as get_alpha_vantage_global_news,
)
from .alpha_vantage_common import AlphaVantageRateLimitError
from .polaris import (
get_stock_data as get_polaris_stock,
get_indicators as get_polaris_indicators,
get_fundamentals as get_polaris_fundamentals,
get_balance_sheet as get_polaris_balance_sheet,
get_cashflow as get_polaris_cashflow,
get_income_statement as get_polaris_income_statement,
get_insider_transactions as get_polaris_insider_transactions,
get_news as get_polaris_news,
get_global_news as get_polaris_global_news,
get_sentiment_score as get_polaris_sentiment_score,
get_sector_analysis as get_polaris_sector_analysis,
get_news_impact as get_polaris_news_impact,
)
# Configuration and routing logic
from .config import get_config
@ -57,12 +71,21 @@ TOOLS_CATEGORIES = {
"get_global_news",
"get_insider_transactions",
]
},
"sentiment_analysis": {
"description": "Sentiment scoring, trading signals, and news impact (Polaris-exclusive)",
"tools": [
"get_sentiment_score",
"get_sector_analysis",
"get_news_impact",
]
}
}
VENDOR_LIST = [
"yfinance",
"alpha_vantage",
"polaris",
]
# Mapping of methods to their vendor-specific implementations
@ -71,41 +94,60 @@ VENDOR_METHODS = {
"get_stock_data": {
"alpha_vantage": get_alpha_vantage_stock,
"yfinance": get_YFin_data_online,
"polaris": get_polaris_stock,
},
# technical_indicators
"get_indicators": {
"alpha_vantage": get_alpha_vantage_indicator,
"yfinance": get_stock_stats_indicators_window,
"polaris": get_polaris_indicators,
},
# fundamental_data
"get_fundamentals": {
"alpha_vantage": get_alpha_vantage_fundamentals,
"yfinance": get_yfinance_fundamentals,
"polaris": get_polaris_fundamentals,
},
"get_balance_sheet": {
"alpha_vantage": get_alpha_vantage_balance_sheet,
"yfinance": get_yfinance_balance_sheet,
"polaris": get_polaris_balance_sheet,
},
"get_cashflow": {
"alpha_vantage": get_alpha_vantage_cashflow,
"yfinance": get_yfinance_cashflow,
"polaris": get_polaris_cashflow,
},
"get_income_statement": {
"alpha_vantage": get_alpha_vantage_income_statement,
"yfinance": get_yfinance_income_statement,
"polaris": get_polaris_income_statement,
},
# news_data
"get_news": {
"alpha_vantage": get_alpha_vantage_news,
"yfinance": get_news_yfinance,
"polaris": get_polaris_news,
},
"get_global_news": {
"yfinance": get_global_news_yfinance,
"alpha_vantage": get_alpha_vantage_global_news,
"polaris": get_polaris_global_news,
},
"get_insider_transactions": {
"alpha_vantage": get_alpha_vantage_insider_transactions,
"yfinance": get_yfinance_insider_transactions,
"polaris": get_polaris_insider_transactions,
},
# sentiment_analysis (Polaris-exclusive)
"get_sentiment_score": {
"polaris": get_polaris_sentiment_score,
},
"get_sector_analysis": {
"polaris": get_polaris_sector_analysis,
},
"get_news_impact": {
"polaris": get_polaris_news_impact,
},
}

582
tradingagents/dataflows/polaris.py Normal file
View File

@ -0,0 +1,582 @@
"""
Polaris Knowledge API data vendor for TradingAgents.
Polaris provides sentiment-scored intelligence briefs, composite trading signals,
technical indicators, financial data, and news impact analysis. Unlike raw data
feeds, every Polaris response includes confidence scores, bias analysis, and
NLP-derived metadata that enriches agent decision-making.
Setup:
pip install polaris-news
export POLARIS_API_KEY=pr_live_xxx # Free: 1,000 credits/month at thepolarisreport.com
API docs: https://thepolarisreport.com/api-reference
"""
import os
import threading
from typing import Annotated
from datetime import datetime
try:
from cachetools import TTLCache
except ImportError:
# Fallback if cachetools not installed
from functools import lru_cache
TTLCache = None
# ---------------------------------------------------------------------------
# Configuration
# ---------------------------------------------------------------------------
_CACHE_TTL = 300 # 5 minutes
_CACHE_MAX = 500
# Thread-safe TTL cache (preferred) with fallback to simple dict
if TTLCache is not None:
_cache = TTLCache(maxsize=_CACHE_MAX, ttl=_CACHE_TTL)
_cache_lock = threading.Lock()
else:
_cache = {}
_cache_lock = threading.Lock()
_client_instance = None
_client_lock = threading.Lock()
def _get_client():
"""Lazy-initialize Polaris client (thread-safe singleton)."""
global _client_instance
if _client_instance is not None:
return _client_instance
with _client_lock:
if _client_instance is not None:
return _client_instance
try:
from polaris_news import PolarisClient
except ImportError:
raise ImportError(
"polaris-news is required for the Polaris data vendor. "
"Install it with: pip install polaris-news"
)
api_key = os.environ.get("POLARIS_API_KEY", "demo")
_client_instance = PolarisClient(api_key=api_key)
return _client_instance
def _cached(key: str):
"""Check cache for a key. Returns cached value or None (thread-safe)."""
with _cache_lock:
return _cache.get(key)
def _set_cache(key: str, data: str):
"""Store data in cache (thread-safe)."""
with _cache_lock:
_cache[key] = data
# ---------------------------------------------------------------------------
# Core Stock APIs
# ---------------------------------------------------------------------------
def get_stock_data(
symbol: Annotated[str, "ticker symbol of the company"],
start_date: Annotated[str, "Start date in yyyy-mm-dd format"],
end_date: Annotated[str, "End date in yyyy-mm-dd format"],
) -> str:
"""Fetch OHLCV stock data from Polaris (via multi-provider: Yahoo/TwelveData/FMP)."""
cache_key = f"stock:{symbol}:{start_date}:{end_date}"
cached = _cached(cache_key)
if cached:
return cached
client = _get_client()
# Determine range from date span
start = datetime.strptime(start_date, "%Y-%m-%d")
end = datetime.strptime(end_date, "%Y-%m-%d")
days = (end - start).days
if days <= 30:
range_param = "1mo"
elif days <= 90:
range_param = "3mo"
elif days <= 180:
range_param = "6mo"
elif days <= 365:
range_param = "1y"
elif days <= 730:
range_param = "2y"
else:
range_param = "5y"
try:
data = client.candles(symbol, interval="1d", range=range_param)
except Exception as e:
return f"Error fetching stock data for {symbol}: {e}"
candles = data.get("candles", [])
if not candles:
return f"No data found for symbol '{symbol}' between {start_date} and {end_date}"
# Filter to requested date range
candles = [c for c in candles if start_date <= c["date"] <= end_date]
# Format as CSV (matching yfinance output format)
header = f"# Stock data for {symbol.upper()} from {start_date} to {end_date}\n"
header += f"# Source: Polaris Knowledge API (multi-provider: Yahoo/TwelveData/FMP)\n"
header += f"# Total records: {len(candles)}\n\n"
csv = "Date,Open,High,Low,Close,Volume\n"
for c in candles:
csv += f"{c['date']},{c['open']},{c['high']},{c['low']},{c['close']},{c['volume']}\n"
result = header + csv
_set_cache(cache_key, result)
return result
# ---------------------------------------------------------------------------
# Technical Indicators
# ---------------------------------------------------------------------------
def get_indicators(
symbol: Annotated[str, "ticker symbol of the company"],
indicator: Annotated[str, "technical indicator to get"],
curr_date: Annotated[str, "Current trading date, YYYY-mm-dd"],
look_back_days: Annotated[int, "how many days to look back"],
) -> str:
"""Fetch technical indicators from Polaris (20 indicators + signal summary)."""
cache_key = f"indicators:{symbol}:{indicator}:{curr_date}:{look_back_days}"
cached = _cached(cache_key)
if cached:
return cached
client = _get_client()
# Map common indicator names to Polaris types
indicator_map = {
"close_50_sma": "sma", "close_20_sma": "sma", "close_200_sma": "sma",
"rsi_14": "rsi", "rsi": "rsi",
"macd": "macd", "macds": "macd", "macdh": "macd",
"boll": "bollinger", "boll_ub": "bollinger", "boll_lb": "bollinger",
"atr": "atr", "atr_14": "atr",
"stoch": "stochastic", "stochrsi": "stochastic",
"adx": "adx", "williams_r": "williams_r",
"cci": "cci", "mfi": "mfi", "roc": "roc",
"obv": "obv", "vwap": "vwap",
}
polaris_type = indicator_map.get(indicator.lower(), indicator.lower())
# Determine range
if look_back_days <= 30:
range_param = "1mo"
elif look_back_days <= 90:
range_param = "3mo"
elif look_back_days <= 180:
range_param = "6mo"
else:
range_param = "1y"
# Try specific indicator first, fall back to full technicals
try:
if polaris_type in ["sma", "ema", "rsi", "macd", "bollinger", "atr",
"stochastic", "adx", "obv", "vwap", "williams_r",
"cci", "mfi", "roc", "ppo", "trix", "donchian",
"parabolic_sar", "ichimoku", "fibonacci"]:
data = client.indicators(symbol, type=polaris_type, range=range_param)
else:
data = client.technicals(symbol, range=range_param)
except Exception as e:
return f"Error fetching indicators for {symbol}: {e}"
values = data.get("values", [])
header = f"# Technical Indicator: {indicator} for {symbol.upper()}\n"
header += f"# Source: Polaris Knowledge API\n"
header += f"# Period: {range_param} | Data points: {len(values)}\n\n"
if isinstance(values, list) and values:
# Format based on indicator type
first = values[0]
if "value" in first:
csv = "Date,Value\n"
for v in values:
csv += f"{v['date']},{v['value']}\n"
elif "macd" in first:
csv = "Date,MACD,Signal,Histogram\n"
for v in values:
csv += f"{v['date']},{v.get('macd','')},{v.get('signal','')},{v.get('histogram','')}\n"
elif "upper" in first:
csv = "Date,Upper,Middle,Lower\n"
for v in values:
csv += f"{v['date']},{v.get('upper','')},{v.get('middle','')},{v.get('lower','')}\n"
elif "k" in first:
csv = "Date,K,D\n"
for v in values:
csv += f"{v['date']},{v.get('k','')},{v.get('d','')}\n"
else:
csv = str(values)
elif isinstance(values, dict):
# Fibonacci or similar
csv = str(values)
else:
csv = "No indicator data available"
result = header + csv
_set_cache(cache_key, result)
return result
# ---------------------------------------------------------------------------
# Fundamental Data
# ---------------------------------------------------------------------------
def _get_financials_cached(symbol: str) -> dict:
"""Shared cached financials fetch — used by fundamentals, balance_sheet, cashflow, income_statement."""
cache_key = f"financials_raw:{symbol}"
cached = _cached(cache_key)
if cached:
return cached
client = _get_client()
data = client.financials(symbol)
_set_cache(cache_key, data)
return data
def get_fundamentals(
symbol: Annotated[str, "ticker symbol of the company"],
) -> str:
"""Fetch company fundamentals from Polaris (via Yahoo Finance quoteSummary)."""
cache_key = f"fundamentals:{symbol}"
cached = _cached(cache_key)
if cached:
return cached
try:
data = _get_financials_cached(symbol)
except Exception as e:
return f"Error fetching fundamentals for {symbol}: {e}"
result = f"# Company Fundamentals: {data.get('company_name', symbol)}\n"
result += f"# Source: Polaris Knowledge API\n\n"
result += f"Sector: {data.get('sector', 'N/A')}\n"
result += f"Industry: {data.get('industry', 'N/A')}\n"
result += f"Market Cap: {data.get('market_cap_formatted', 'N/A')}\n"
result += f"P/E Ratio: {data.get('pe_ratio', 'N/A')}\n"
result += f"Forward P/E: {data.get('forward_pe', 'N/A')}\n"
result += f"EPS: {data.get('eps', 'N/A')}\n"
result += f"Revenue: {data.get('revenue_formatted', 'N/A')}\n"
result += f"EBITDA: {data.get('ebitda_formatted', 'N/A')}\n"
result += f"Profit Margin: {data.get('profit_margin', 'N/A')}\n"
result += f"Debt/Equity: {data.get('debt_to_equity', 'N/A')}\n"
result += f"ROE: {data.get('return_on_equity', 'N/A')}\n"
result += f"Beta: {data.get('beta', 'N/A')}\n"
result += f"52-Week High: {data.get('fifty_two_week_high', 'N/A')}\n"
result += f"52-Week Low: {data.get('fifty_two_week_low', 'N/A')}\n"
_set_cache(cache_key, result)
return result
def get_balance_sheet(
symbol: Annotated[str, "ticker symbol of the company"],
) -> str:
"""Fetch balance sheet from Polaris."""
try:
data = _get_financials_cached(symbol)
except Exception as e:
return f"Error fetching balance sheet for {symbol}: {e}"
sheets = data.get("balance_sheets", [])
result = f"# Balance Sheet: {symbol.upper()}\n# Source: Polaris Knowledge API\n\n"
result += "Date,Total Assets,Total Liabilities,Total Equity\n"
for s in sheets:
result += f"{s['date']},{s['total_assets']},{s['total_liabilities']},{s['total_equity']}\n"
return result
def get_cashflow(
symbol: Annotated[str, "ticker symbol of the company"],
) -> str:
"""Fetch cash flow data from Polaris."""
try:
data = _get_financials_cached(symbol)
except Exception as e:
return f"Error fetching cashflow for {symbol}: {e}"
result = f"# Cash Flow: {symbol.upper()}\n# Source: Polaris Knowledge API\n\n"
result += f"Free Cash Flow: {data.get('free_cash_flow', 'N/A')}\n"
return result
def get_income_statement(
symbol: Annotated[str, "ticker symbol of the company"],
) -> str:
"""Fetch income statement from Polaris."""
try:
data = _get_financials_cached(symbol)
except Exception as e:
return f"Error fetching income statement for {symbol}: {e}"
stmts = data.get("income_statements", [])
result = f"# Income Statement: {symbol.upper()}\n# Source: Polaris Knowledge API\n\n"
result += "Date,Revenue,Net Income,Gross Profit\n"
for s in stmts:
result += f"{s['date']},{s['revenue']},{s['net_income']},{s['gross_profit']}\n"
return result
# ---------------------------------------------------------------------------
# News & Intelligence (Polaris advantage — sentiment-scored, not raw headlines)
# ---------------------------------------------------------------------------
def get_news(
symbol: Annotated[str, "ticker symbol of the company"],
start_date: Annotated[str, "Start date in yyyy-mm-dd format"],
end_date: Annotated[str, "End date in yyyy-mm-dd format"],
) -> str:
"""Fetch sentiment-scored intelligence briefs from Polaris.
Unlike raw news feeds, each brief includes:
- Confidence score (0-1)
- Bias score and direction
- Counter-arguments
- Entity-level sentiment (-1.0 to +1.0)
"""
cache_key = f"news:{symbol}:{start_date}:{end_date}"
cached = _cached(cache_key)
if cached:
return cached
client = _get_client()
try:
data = client.search(symbol, per_page=20)
# Handle both dict and typed response objects
if hasattr(data, '__dict__') and not isinstance(data, dict):
data = data.__dict__ if hasattr(data, '__dict__') else {}
if isinstance(data, dict):
briefs = data.get("briefs", [])
else:
briefs = getattr(data, 'briefs', [])
except Exception as e:
return f"Error fetching news for {symbol}: {e}"
if not briefs:
return f"No intelligence briefs found for {symbol}"
result = f"# Intelligence Briefs for {symbol.upper()}\n"
result += f"# Source: Polaris Knowledge API (sentiment-scored, bias-analyzed)\n"
result += f"# Total: {len(briefs)} briefs\n\n"
def _get(obj, key, default='N/A'):
"""Get attribute from dict or object."""
if isinstance(obj, dict):
return obj.get(key, default)
return getattr(obj, key, default)
for b in briefs:
prov = _get(b, "provenance", {})
result += f"--- Brief: {_get(b, 'id', '')} ---\n"
result += f"Date: {_get(b, 'published_at', '')}\n"
result += f"Headline: {_get(b, 'headline', '')}\n"
result += f"Summary: {_get(b, 'summary', '')}\n"
result += f"Category: {_get(b, 'category', '')}\n"
result += f"Confidence: {_get(prov, 'confidence_score', 'N/A')}\n"
result += f"Bias Score: {_get(prov, 'bias_score', 'N/A')}\n"
result += f"Review Status: {_get(prov, 'review_status', 'N/A')}\n"
result += f"Sentiment: {_get(b, 'sentiment', 'N/A')}\n"
result += f"Impact Score: {_get(b, 'impact_score', 'N/A')}\n"
entities = _get(b, "entities_enriched", []) or []
if entities:
ent_str = ", ".join(
f"{_get(e, 'name', '?')}({_get(e, 'sentiment_score', '?')})"
for e in (entities[:5] if isinstance(entities, list) else [])
)
result += f"Entities: {ent_str}\n"
ca = _get(b, "counter_argument", None)
if ca:
result += f"Counter-Argument: {str(ca)[:200]}...\n"
result += "\n"
_set_cache(cache_key, result)
return result
def get_global_news(
start_date: Annotated[str, "Start date in yyyy-mm-dd format"],
end_date: Annotated[str, "End date in yyyy-mm-dd format"],
) -> str:
"""Fetch global intelligence feed from Polaris with sentiment and bias scoring."""
cache_key = f"global_news:{start_date}:{end_date}"
cached = _cached(cache_key)
if cached:
return cached
client = _get_client()
try:
data = client.feed(per_page=20)
if hasattr(data, '__dict__') and not isinstance(data, dict):
data = data.__dict__ if hasattr(data, '__dict__') else {}
if isinstance(data, dict):
briefs = data.get("briefs", [])
else:
briefs = getattr(data, 'briefs', [])
except Exception as e:
return f"Error fetching global news: {e}"
result = f"# Global Intelligence Feed\n"
result += f"# Source: Polaris Knowledge API\n"
result += f"# Briefs: {len(briefs)}\n\n"
def _get2(obj, key, default='N/A'):
if isinstance(obj, dict):
return obj.get(key, default)
return getattr(obj, key, default)
for b in briefs:
prov = _get2(b, "provenance", {})
pub = str(_get2(b, 'published_at', ''))[:10]
result += f"[{pub}] [{_get2(b, 'category', '')}] "
result += f"{_get2(b, 'headline', '')} "
result += f"(confidence={_get2(prov, 'confidence_score', '?')}, "
result += f"bias={_get2(prov, 'bias_score', '?')}, "
result += f"sentiment={_get2(b, 'sentiment', '?')})\n"
_set_cache(cache_key, result)
return result
def get_insider_transactions(
symbol: Annotated[str, "ticker symbol of the company"],
) -> str:
"""Fetch SEC EDGAR earnings filings via Polaris."""
client = _get_client()
try:
data = client.transcripts(symbol, days=365)
except Exception as e:
return f"Error fetching filings for {symbol}: {e}"
filings = data.get("filings", [])
result = f"# SEC Filings for {symbol.upper()}\n"
result += f"# Source: Polaris Knowledge API (SEC EDGAR)\n\n"
result += "Date,Form,Description,URL\n"
for f in filings[:20]:
result += f"{f.get('date', '')},{f.get('form', '')},{f.get('description', '')},{f.get('filing_url', '')}\n"
return result
# ---------------------------------------------------------------------------
# Polaris-Exclusive: Sentiment & Trading Signals
# (Not available from Yahoo Finance or Alpha Vantage)
# ---------------------------------------------------------------------------
def get_sentiment_score(
symbol: Annotated[str, "ticker symbol of the company"],
) -> str:
"""Get composite trading signal from Polaris.
Returns a multi-factor score combining:
- Sentiment (40% weight)
- Momentum (25% weight)
- Coverage velocity (20% weight)
- Event proximity (15% weight)
Not available from any other data vendor.
"""
cache_key = f"sentiment:{symbol}"
cached = _cached(cache_key)
if cached:
return cached
client = _get_client()
try:
data = client.ticker_score(symbol)
except Exception as e:
return f"Error fetching sentiment score for {symbol}: {e}"
result = f"# Composite Trading Signal: {symbol.upper()}\n"
result += f"# Source: Polaris Knowledge API (exclusive)\n\n"
result += f"Signal: {data.get('signal', 'N/A')}\n"
result += f"Composite Score: {data.get('composite_score', 'N/A')}\n\n"
components = data.get("components", {})
sent = components.get("sentiment", {})
result += f"Sentiment (40%): current_24h={sent.get('current_24h')}, week_avg={sent.get('week_avg')}\n"
mom = components.get("momentum", {})
result += f"Momentum (25%): {mom.get('direction', 'N/A')} (value={mom.get('value')})\n"
vol = components.get("volume", {})
result += f"Volume (20%): {vol.get('briefs_24h')} briefs/24h, velocity={vol.get('velocity_change_pct')}%\n"
evt = components.get("events", {})
result += f"Events (15%): {evt.get('count_7d')} events, latest={evt.get('latest_type')}\n"
_set_cache(cache_key, result)
return result
def get_sector_analysis(
symbol: Annotated[str, "ticker symbol of the company"],
) -> str:
"""Get competitor intelligence for a ticker — same-sector peers with live data."""
cache_key = f"competitors:{symbol}"
cached = _cached(cache_key)
if cached:
return cached
client = _get_client()
try:
data = client.competitors(symbol)
except Exception as e:
return f"Error fetching sector analysis for {symbol}: {e}"
result = f"# Competitor Analysis: {symbol.upper()} ({data.get('sector', 'N/A')})\n"
result += f"# Source: Polaris Knowledge API (exclusive)\n\n"
result += "Ticker,Name,Price,RSI,Sentiment_7d,Briefs_7d\n"
for c in data.get("competitors", []):
result += f"{c.get('ticker')},{c.get('entity_name')},{c.get('price')},{c.get('rsi_14')},{c.get('sentiment_7d')},{c.get('briefs_7d')}\n"
_set_cache(cache_key, result)
return result
def get_news_impact(
symbol: Annotated[str, "ticker symbol of the company"],
) -> str:
"""Measure how news moved the stock price — brief-to-price causation analysis."""
cache_key = f"impact:{symbol}"
cached = _cached(cache_key)
if cached:
return cached
client = _get_client()
try:
data = client.news_impact(symbol)
except Exception as e:
return f"Error fetching news impact for {symbol}: {e}"
result = f"# News Impact Analysis: {symbol.upper()}\n"
result += f"# Source: Polaris Knowledge API (exclusive)\n\n"
result += f"Briefs Analyzed: {data.get('briefs_analyzed', 0)}\n"
result += f"Avg 1-Day Impact: {data.get('avg_impact_1d_pct', 'N/A')}%\n"
result += f"Avg 3-Day Impact: {data.get('avg_impact_3d_pct', 'N/A')}%\n\n"
best = data.get("best_impact", {})
if best:
result += f"Best Impact: {best.get('headline', '')[:60]} (+{best.get('impact_1d_pct')}%)\n"
worst = data.get("worst_impact", {})
if worst:
result += f"Worst Impact: {worst.get('headline', '')[:60]} ({worst.get('impact_1d_pct')}%)\n"
_set_cache(cache_key, result)
return result

10
tradingagents/default_config.py
View File

@ -23,13 +23,15 @@ DEFAULT_CONFIG = {
# Data vendor configuration
# Category-level configuration (default for all tools in category)
"data_vendors": {
"core_stock_apis": "yfinance", # Options: alpha_vantage, yfinance
"technical_indicators": "yfinance", # Options: alpha_vantage, yfinance
"fundamental_data": "yfinance", # Options: alpha_vantage, yfinance
"news_data": "yfinance", # Options: alpha_vantage, yfinance
"core_stock_apis": "yfinance", # Options: alpha_vantage, yfinance, polaris
"technical_indicators": "yfinance", # Options: alpha_vantage, yfinance, polaris
"fundamental_data": "yfinance", # Options: alpha_vantage, yfinance, polaris
"news_data": "yfinance", # Options: alpha_vantage, yfinance, polaris
"sentiment_analysis": "polaris", # Polaris-exclusive: trading signals, sector analysis, news impact
},
# Tool-level configuration (takes precedence over category-level)
"tool_vendors": {
# Example: "get_stock_data": "alpha_vantage", # Override category default
# Example: "get_news": "polaris", # Use Polaris for sentiment-scored news
},
}