The AI Coach
The coach is a post-trade review assistant that reads the context you've already captured — your trades, notes, tags, charts, and rules — and helps summarize what worked, what didn't, and what to focus on next. The aim is feedback that's specific to you rather than generic trading advice.
Codify your edge first
The coach works best against a standard you define. The home for that is a setup workspace in Settings — the first pieces (connecting a model provider and editing your playbook) are there today, and the rest is filling in:
- add your own language-model API key (OpenAI works today; other providers are planned),
- complete a short trader self-assessment (coming),
- define your rules and setups, and
- tune how the coach reviews (coming).
That rubric is what turns "was this a good trade?" into "did this match my A+ setup?"
How a review works
With a model key configured, asking the coach to review a day will:
- read your trades against your playbook and criteria,
- point to where you followed your plan and where you drifted, and
- draft a recap and feedback in your voice.
You always stay in control
This principle holds throughout the design:
- The coach proposes; it never posts on its own.
- Every recap and note arrives as a draft you read and edit.
- Every number in a review comes from deterministic calculations over your actual trades — the model explains the math, it's never allowed to invent it.
- Nothing is sent anywhere unless you ask for a review. When you do, the trades, notes, and tags for that session go to the model provider you configured, under your own API key — never through our servers, and never automatically in the background. No key configured means nothing ever leaves your machine.
What you can do today
While the coach matures, the most valuable thing is to build the habit it depends on: import your trades, write daily recaps, and tag consistently. That context is exactly what the coach reads when you ask for a review — follow the Roadmap to see the rest come together.