Two bugs were causing the signal card HTML to display as literal text:
1. CommonMark HTML block termination: Streamlit's markdown parser
(CommonMark-compliant) terminates a <div> block at the first blank
line. Empty optional fields (name_html, desc_html, risk_badge_html)
left whitespace-only lines in the f-string template, ending the HTML
block and causing the parser to render subsequent tags as text.
Fixed by building HTML from a parts list and only appending optional
elements when non-empty — no blank lines can appear in output.
2. Unescaped HTML chars in LLM-generated text: reason fields from the
ranker contained raw > and & characters (e.g. '>5% move',
'50 & 200 SMA') that corrupted the HTML structure. Fixed by running
all LLM-generated fields through html.escape() before interpolation.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add daily price movement display with color coding (green/red)
- Add 1D (intraday) and 7D chart options with granular data:
- 1D: 5-minute interval for detailed intraday view
- 7D: hourly interval for smooth 7-day chart
- Fix discontinuous chart rendering by plotting against sequential index for intraday data
- Eliminate overnight/weekend gaps in hourly charts
- Add timezone normalization for consistent date handling between daily and intraday data
- Improve fallback logic when data is sparse
- Better handling of yfinance column names (Datetime vs Date)
Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
- Add GitHub Actions workflow for daily discovery (8:30 AM ET, weekdays)
- Add headless run_daily_discovery.py script for scheduling
- Expand options_flow scanner to use tickers.txt with parallel execution
- Add recommendation history section to Performance page with filters and charts
- Fix strategy name normalization (momentum/Momentum/Momentum-Hype → momentum)
- Fix strategy metrics to count all recs, not just evaluated ones
- Add error handling to Streamlit page rendering
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add ML signal scanner results table logging
- Add log_prompts_console config flag for prompt visibility control
- Expand ranker investment thesis to 4-6 sentence structured reasoning
- Linter auto-formatting across modified files
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Major additions:
- ML win probability scanner: scans ticker universe using trained
LightGBM/TabPFN model, surfaces candidates with P(WIN) above threshold
- 30-feature engineering pipeline (20 base + 10 interaction features)
computed from OHLCV data via stockstats + pandas
- Triple-barrier labeling for training data generation
- Dataset builder and training script with calibration analysis
- Discovery enrichment: confluence scoring, short interest extraction,
earnings estimates, options signal normalization, quant pre-score
- Configurable prompt logging (log_prompts_console flag)
- Enhanced ranker investment thesis (4-6 sentence reasoning)
- Typed DiscoveryConfig dataclass for all discovery settings
- Console price charts for visual ticker analysis
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Youssef Aitousarrah