> ## Documentation Index
> Fetch the complete documentation index at: https://docs.worthcall.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Daily Routine

> Pre-market prep, regular session execution, and EOD review.

# Daily Trading Routine

## Pre-market (4:00 AM – 9:30 AM ET)

<Steps>
  <Step title="Check status">
    Navigate to `/status` or run `get_status` to confirm TWS is connected and market session is "pre-market".
  </Step>

  <Step title="Review overnight signals">
    Check `/signals` for any Holly alerts that were auto-evaluated overnight. Look for scores above 70.
  </Step>

  <Step title="Build watchlist">
    Use the screener at `/screener` with `day_gainers` and `most_actives` from prior session. Calculate gap % manually from current quotes vs prior close.
  </Step>

  <Step title="Check analytics">
    Open `/analytics` to remind yourself of your rules:

    * Skip 9 AM entries (your biggest leak)
    * Lean into shorts (your edge)
    * Filter Holly Grail over Neo
  </Step>
</Steps>

## Regular session (9:30 AM – 4:00 PM ET)

<Steps>
  <Step title="Wait for 10 AM">
    Your data shows 9 AM is a consistent leak. Wait for the first 30 minutes to settle.
  </Step>

  <Step title="Evaluate setups">
    Use the ensemble evaluation (MCP `eval_stats` or `/evals`) to score any setup through all 3 models. Only trade scores above your threshold.
  </Step>

  <Step title="Size and execute">
    Use `size_position` to calculate shares from entry/stop. Execute with `place_advanced_bracket` for full bracket protection.
  </Step>

  <Step title="Monitor positions">
    `/orders` shows open orders. `/account` shows positions with portfolio exposure. The exit plan tracks your TP ladder and trailing stop.
  </Step>
</Steps>

## Post-market (4:00 PM – 8:00 PM ET)

<Steps>
  <Step title="Review executions">
    `/executions` shows today's fills with realized P\&L and commissions.
  </Step>

  <Step title="Journal trades">
    Add entries at `/journal/new` with reasoning, setup type, and outcome tags. This feeds the feature attribution model.
  </Step>

  <Step title="Record outcomes">
    Link trade results to evaluations with `record_outcome` so the ensemble can learn from your actual trades.
  </Step>

  <Step title="Check session state">
    `/session` shows daily P\&L, trade count, and whether you hit any guardrails.
  </Step>
</Steps>

## Weekend

* Review `/analytics` for weekly trends
* Check `/drift` for model accuracy degradation
* Run `edge_report` to validate ensemble edge is real (walk-forward validation)
* Consider `tune_risk_params` to auto-calibrate position sizing from recent outcomes