Playbooks
Save the setup, invalidation, risk, and market rules every review should use.
A process journal for discretionary traders
Turn chart evidence into a repeatable trading process.
Capture the plan before hindsight changes it, compare every setup with your saved rules, then learn from outcomes inside the exact strategy or backtest that produced them.
Evidence, not predictionsPrivate account workspaceBrowser OAuth
New trade reviewSanitised demo data
01Editable before save
02Scoped before analysis
03Outcome before insight
04Rules before AI opinion
The complete journal
Every part of the review cycle has a job.
No single total can explain a trader testing several instruments and strategies. The journal preserves the hierarchy from playbook to session to individual trade, then carries that scope into dashboards and reports.
Backtest sessions
Keep each hypothesis, instrument, timeframe, and test period in its own evidence set.
Trade capture
Turn a TradingView link or chart image into an editable evidence preview before save.
Outcome review
Add follow-up evidence, close the trade, and compare the original plan with what happened.
Scoped dashboards
Filter metrics by playbook, backtest, instrument, and timeframe without blending unrelated trades.
Evidence reports
Review timing, process, and outcome themes with visible sample-size cautions.
A disciplined loop
From saved rules to the next useful experiment.
The product order follows the real decision flow. Each step makes the next one more reliable.
- 01
Define the plan
Create a playbook, then start a bounded backtest session when you are testing a hypothesis.
- 02
Capture the evidence
Paste a TradingView link or add an image. Verify every core field and its confidence before saving.
- 03
Close the loop
Post outcome evidence with its date and timezone so the trade closes as a win, loss, or breakeven.
- 04
Review one cohort
Choose the playbook or session you want to understand, then turn repeated themes into the next test.
Clean cohort analysis
See the dashboard for the strategy you are actually reviewing.
Start with the whole journal, then narrow to one playbook, one backtest, one instrument, or one timeframe. The active scope stays visible beside the sample size so mixed evidence is never mistaken for one system.
- Cumulative performance and process metrics use the same cohort.
- The review queue shows evidence and strategy-fit status.
- Scope coaching turns supported themes into the next test.
Scoped dashboardPlaybook + backtest
AI with boundaries
A second set of eyes on your process—not a market oracle.
What the review uses
- Your saved playbook rules and required evidence
- The original chart, extracted fields, provenance, and confidence
- Follow-up evidence, timestamps, and the recorded outcome
- The selected report cohort and its sample size
What it is designed to return
- Setup-fit checks and missing evidence
- Plan-versus-outcome feedback
- Repeated process strengths and drags
- Ideas for a bounded next backtest
What it will not do
- Predict profit or guarantee an outcome
- Issue buy or sell instructions
- Invent unreadable chart facts
- Hide degraded mode when AI is offline
Evidence reportTiming + process
Reports that stay honest
Find when a strategy behaves differently—and whether the sample supports it.
Trade and follow-up timestamps let reports compare weekdays, recorded local time blocks, market sessions, outcome timing, and holding periods. Themes stay tied to the chosen playbook or backtest.
Timing evidenceEntry, outcome, session, holding periodNext experimentOne testable process change
Browser connection
Your journal account and ChatGPT connection remain separate.
Sign in to the journal normally, then authorize your own ChatGPT account on the official website with a one-time device code. There is nothing to install, and the journal remains usable in deterministic degraded mode when AI is unavailable.
Before you start
Common questions.
The detailed setup guide, field explanations, best practices, and troubleshooting live inside every signed-in workspace.
Does the journal place trades or tell me what to buy?
No. It reviews evidence against the rules you saved. It does not execute orders, predict profit, or issue buy and sell instructions.
Can I use it without the AI connection?
Yes. Manual editing, deterministic extraction, risk/reward, and saved-rule checks remain available. The interface states clearly when AI review is unavailable.
Why separate playbooks and backtest sessions?
A playbook is the reusable strategy. A backtest session is one bounded experiment using that strategy. Keeping both prevents unrelated tests from being blended in dashboards and reports.
Can I connect ChatGPT from a phone?
Yes. The responsive website starts a device-code OAuth flow that completes on the official ChatGPT website in your browser; no installed app folder or terminal is required.
What data should I record for useful reports?
Verify the core levels, choose the correct playbook or session, record the chart timezone and trade time, then add follow-up evidence and its timestamp. Consistent evidence creates meaningful cohorts.
Build the evidence
Make every review part of the next decision.
Start with one playbook, one bounded test, and one verified trade.