LLM Trading Workflow

Streamlined process: Get candidates → Ask LLM → Validate response → Execute trades

Workflow Steps

1
Load Candidates
Latest screener results
2
Copy Prompt & Payload
Send to your LLM
3
Validate Response
Check constraints
4
Execute & Record
Trade & log fills

Holdings Export

Generate holdings.csv and crypto_holdings.csv from your current portfolio for use with Python screeners.

Import Python Candidates

Import llm_candidates.json and llm_candidates_crypto.json from python/tfsa-llm/ to make them available in the web app.

Python Screener Commands

Stocks
cd python/scripts && python screener_top40_fractional.py --cash 11 --fractional --min-trade-size 1 --holdings ../outputs/holdings.csv --tsx-only
Crypto
cd python/scripts && python screener_crypto_top40_fractional.py --cash 11 --fractional --min-trade-size 1 --holdings ../outputs/crypto_holdings.csv
📁 Next Step
After running the screeners, use the green “Import Candidates” button above to make them available in the web app.
💡 Run these commands from your project root directory (where holdings.csv is located)

No stocks candidates loaded