From workflows
Access financial data from LSEG/Refinitiv via the lseg.data Python API. Supports fundamentals, market data, ESG, symbology, deals, loans, funds, screening, and news with strict validation.
How this skill is triggered — by the user, by Claude, or both
Slash command
/workflows:lseg-dataThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
- [Query Enforcement](#query-enforcement)
examples/fundamentals_query.pyexamples/historical_pricing.ipynbexamples/stock_screener.ipynbreferences/api-discovery.mdreferences/corporate-governance.mdreferences/equity-new-issues.mdreferences/esg.mdreferences/fscreen.mdreferences/fund-details.mdreferences/fundamentals.mdreferences/infrastructure.mdreferences/joint-ventures.mdreferences/mna.mdreferences/municipal-bonds.mdreferences/news.mdreferences/pricing.mdreferences/private-equity.mdreferences/screening.mdreferences/symbology.mdreferences/syndicated-loans.mdAccess financial data from LSEG (London Stock Exchange Group), formerly Refinitiv, via the lseg.data Python library.
Before claiming ANY LSEG query succeeded, follow these steps:
.head() or .sample()This is not negotiable. Skipping result inspection is NOT HELPFUL — the user builds analysis on data with undetected quality problems.
.O, .N, .L, .T) before querying.get_data() 10,000 data points, get_history() 3,000 rows) — many small queries still hit the session cap. Batch instead of looping..head() or .sample() inspection → STOP. Handing over uninspected data gives the user undetected quality problems — unhelpful on its own terms.Before EVERY data retrieval claim, verify the following:
For ld.get_data() (fundamentals/ESG):
.head() or .sample() executedFor ld.get_history() (time series):
For symbol_conversion.Definition() (mapping):
For ALL queries:
open_session() at start, close_session() at endTo get started with LSEG Data Library, initialize a session and execute queries:
import lseg.data as ld
# Initialize session
ld.open_session()
# Get fundamentals
df = ld.get_data(
universe=[‘AAPL.O’, ‘MSFT.O’],
fields=[‘TR.CompanyName’, ‘TR.Revenue’, ‘TR.EPS’]
)
print(df.head()) # Inspect sample data
# Get historical prices
prices = ld.get_history(
universe=’AAPL.O’,
fields=[‘OPEN’, ‘HIGH’, ‘LOW’, ‘CLOSE’, ‘VOLUME’],
start=‘2023-01-01’,
end=‘2023-12-31’
)
print(prices.head()) # Inspect sample data
# Close session
ld.close_session()
Configure LSEG authentication using either a config file or environment variables.
Create lseg-data.config.json:
{
“sessions”: {
“default”: “platform.ldp”,
“platform”: {
“ldp”: {
“app-key”: “YOUR_APP_KEY”,
“username”: “YOUR_MACHINE_ID”,
“password”: “YOUR_PASSWORD”
}
}
}
}
Set the following environment variables for LSEG authentication:
# Configure LSEG credentials via environment variables
export RDP_USERNAME=”YOUR_MACHINE_ID”
export RDP_PASSWORD=”YOUR_PASSWORD”
export RDP_APP_KEY=”YOUR_APP_KEY”
| API | Use Case | Example |
|---|---|---|
ld.get_data() | Point-in-time data | Fundamentals, ESG scores |
ld.get_history() | Time series | Historical prices, OHLCV |
ld.news.get_headlines() | News headlines | Company news, topic filtering |
symbol_conversion.Definition() | ID mapping | RIC ↔ ISIN ↔ CUSIP |
| Prefix | Type | Example |
|---|---|---|
TR. | Refinitiv fields | TR.Revenue, TR.EPS |
TR.MnA | Mergers & Acquisitions | TR.MnAAcquirorName, TR.MnADealValue |
TR.NI | Equity/New Issues (IPOs) | TR.NIIssuer, TR.NIOfferPrice |
TR.JV | Joint Ventures/Alliances | TR.JVDealName, TR.JVStatus |
TR.SACT | Shareholder Activism | TR.SACTLeadDissident |
TR.PP | Poison Pills | TR.PPPillAdoptionDate |
TR.LN | Syndicated Loans | TR.LNTotalFacilityAmount |
TR.PJF | Infrastructure/Project Finance | TR.PJFProjectName |
TR.PEInvest | Private Equity/Venture Capital | TR.PEInvestRoundDate |
TR.Muni | Municipal Bonds | TR.MuniIssuerName |
CF_ | Composite (real-time) | CF_LAST, CF_BID |
| Suffix | Exchange | Example |
|---|---|---|
.O | NASDAQ | AAPL.O |
.N | NYSE | IBM.N |
.L | London | VOD.L |
.T | Tokyo | 7203.T |
| Endpoint | Limit |
|---|---|
get_data() | 10,000 data points/request |
get_history() | 3,000 rows/request |
| Session | 500 requests/minute |
references/fundamentals.md - Financial statement fields, ratios, estimatesreferences/esg.md - ESG scores, pillars, controversiesreferences/symbology.md - RIC/ISIN/CUSIP conversionreferences/pricing.md - Historical prices, real-time datareferences/screening.md - Stock screening with Screener objectreferences/fscreen.md - Fund screening (ETFs, mutual funds) with FSCREEN appreferences/fund-details.md - Fund details and characteristicsreferences/news.md - News headlines, pagination, query syntaxreferences/mna.md - Mergers & acquisitions deals (SDC Platinum, 2,683 fields)references/equity-new-issues.md - IPOs, follow-ons, equity offerings (SDC Platinum, 1,708 fields)references/joint-ventures.md - Joint ventures, strategic alliances (SDC Platinum, 301 fields)references/corporate-governance.md - Shareholder activism, poison pills (SDC Platinum)references/syndicated-loans.md - Syndicated loan deals (SDC Platinum)references/infrastructure.md - Infrastructure/project finance deals (SDC Platinum)references/private-equity.md - Private equity/venture capital investments (SDC Platinum)references/municipal-bonds.md - Municipal bond issuances (SDC Platinum)references/api-discovery.md - Reverse-engineering APIs via CDP network monitoringreferences/troubleshooting.md - Common issues and solutionsreferences/wrds-comparison.md - LSEG vs WRDS data mappingexamples/historical_pricing.ipynb - Historical price retrievalexamples/fundamentals_query.py - Fundamental data patternsexamples/stock_screener.ipynb - Dynamic stock screeningscripts/test_connection.py - Validate LSEG connectivityLSEG API samples at ~/resources/lseg-samples/:
Example.RDPLibrary.Python/ - Core API examplesExamples.DataLibrary.Python.AdvancedUsecases/ - Advanced patternsArticle.DataLibrary.Python.Screener/ - Stock screeningInteractive JupyterLab environment with pre-configured LSEG access:
https://workspace.refinitiv.com/codebook/refinitiv.data library{name=’codebook’})# In Codebook, session opens automatically with Workspace auth
import refinitiv.data as rd
rd.open_session() # Returns session with name=’codebook’
# Query data immediately
df = rd.news.get_headlines(‘R:AAPL.O AND SUGGAC’, count=10)
Note: Codebook uses refinitiv.data (older name) rather than lseg.data. Both APIs are equivalent.
When querying market data, account for current date context and market data lag.
Market data typically has T-1 availability, meaning today’s data becomes available tomorrow. Adjust date ranges accordingly.
Use current date context when querying historical prices:
from datetime import datetime, timedelta
# Get recent market data
end_date = datetime.now()
start_date = end_date - timedelta(days=365)
# Adjust to exclude recent data (T-1 for market data availability)
end_date = end_date - timedelta(days=1)
df = ld.get_history(
universe=”AAPL.O”,
fields=[‘CLOSE’],
start=start_date.strftime(‘%Y-%m-%d’),
end=end_date.strftime(‘%Y-%m-%d’)
)
Remember: Always account for the T-1 lag in market data availability.
npx claudepluginhub edwinhu/workflows --plugin workflowsFetches financial data from the EODHD API — historical prices, fundamentals, options, technical indicators, news, sentiment, macro indicators, corporate events, ESG scores, risk analytics, US Treasury rates, and trading hours.
Designs and manages market data infrastructure for financial trading: real-time/delayed feeds, Level 1/2/3 depth, SIP vs direct feeds, vendor selection (Bloomberg, Refinitiv), licensing, entitlements, ticker plants, and data quality.