Every surface, one place.

The app's sidebar, top to bottom. Each section links by anchor, so you can point at any of these directly. The daily loop is short: set up an account, compose and commit a Playbook, log each trade, then run a brief evening Review.

Dashboard

Your standing read, all of it produced by the practice — nothing here is editable. Up top sits your aim (the trader you're becoming), then a one-sentence hero reading of where you stand, and a slim strip tracing your risk (average heat and edge) across recent sessions. The practice itself shows in three labeled bands of three tiles each:

• The practice — Discipline (the slow reading), your Clean Streak, and your Ritual streak (days logged in a row, bad days included).
• Lately · last 30 days — Rule Adherence (per session), Wait (days sat out cleanly), and Within Planned Risk (how often your heat held inside the planned 1R).
• Losses & recovery · last 90 days — Days Since Bad Loss, Good-Loss Ratio, and Repair-Free days.

Below the bands: your rule-by-rule adherence over the last 30 sessions, the freshest patterns from the evidence engine, a heatmap of your recent weeks (one cell per day, shaded clean down through minor and critical lapses), and any milestones in progress. No capital figure sits on the daily dashboard — the rolling realized read lives on the monthly review. If a week passes with no logged day (away periods excused), the streaks read as paused rather than passing off a stale number as fresh.

A brand-new account meets a short begin-here checklist instead: set up an account, commit a playbook, log a first review.

Playbook

The composer where you turn the fixed method into your standing rulebook. The method itself is shown up top, the same for everyone; below it you choose the method's bounded settings: which instruments you trade (MES, MNQ, ES, NQ), one or more session windows, your stop method, your profit style and — for partials — the trim split (the tranches that sum to 100, the last one the runner), SMT (bonus or required), your risk limits (R in dollars, contract cap, daily loss limit, trades per day), an optional gap-size band per instrument, and which direction(s) you trade. Leave direction on both long and short and it stays unconstrained; commit to a single side and an in-direction rule grades each trade against it.

When the settings are right, you commit and lock. The first commit takes effect immediately. After that, changing any setting is an amendment: it opens a cooldown and goes live at the next session open (at least 24 hours out), never mid-session. A pending amendment can be cancelled before it lands. The friction is the point — the scorer can only grade you against a playbook you can't renegotiate on a losing day. Your rules are derived from this committed playbook; there is no separate place to author them.

Trade (logging)

You log each trade at /review/trade — reachable from the Review page's "+ Trade" action and from the Desk's "Log this trade" button. The form is grouped into four parts:

• Setup — did you wait for the inversion candle close, direction, entry window, confluences present, an optional level/setup label.
• Management — stop held, stayed flat, exit followed the split, TP1 a market order, runner locked to break-even or better.
• Price action — reaction and momentum (awareness only, never graded).
• Risk — stop in points, contracts taken, realized P&L, and the trade's MAE and MFE.

Win or loss is read from your realized P&L — you don't mark an outcome. Per-trade P&L, MAE, and MFE are awareness-only and never scored; the day's numbers roll up from them. Notes and screenshots attach to each trade. The graded conduct is the Setup and Management answers, measured against your committed playbook.

Review

The short evening review. It leads with the day's rolled-up Conduct, read back from the trades you logged — the verdict word, the score, the day's P&L, the entry count, and the biggest winner and loser — followed by the day's trades table and a risk strip (MAE, MFE, max contracts) with two whole-day readings you enter by hand: worst point from open and max intraday drawdown. There is no rule checklist (the conduct is already logged per trade) and no morning follow-through.

The review then asks a handful of day-level questions: what pulled at you? (force the trade, overtrade, prove I'm right — plus an intensity scale), two end-of-day checks your trade log can't see (one system start to finish; walked away without buying a new account), and the written reflection (what you did well, what caused mistakes, the lesson, a strength to repeat, tomorrow's focus). You mark whether you'd take the same trades again (yes / no / partial — partial asks you to split which), a felt mark for proud of your execution, and a repair read for whether trading became about fixing something. On a day you sat out, the page instead asks whether you passed a real setup, with an optional note. Past days can be revised, and a day you never logged can be backfilled.

Cards

Shareable cards rendered from your real record. The card type is the daily review — a full-fidelity rendering of a logged day, showing production data, never a mockup. Share links are tokened, anonymous to your account details, and expire after thirty days. If a link lapses, Renew brings it back on the same URL. Deleting a card removes it for good. Each card shows how many times it has been opened and when it was last viewed. The list groups by week — step between weeks with the arrows in the header.

Sessions

The full log, one row per day, opening into a detail page per date: the day's trades, the derived Conduct verdict and execution score, the felt states, the written reflections, and any notes and media. The log reads one month at a time — step between months with the arrows in the header. Any past day can be revised, and a day you never logged can be back-filled with + Session (which opens the Review for that date); scores recompute on save either way.

Weekly

The week as one object: execution average, clean days, critical count, and the week's reflections side by side. Where the Sunday letter is testimony, the weekly is the ledger.

Monthly

The month in whole: summary line, process and outcome side by side, a calendar heatmap, then the lessons you captured — and beneath them, one written answer to "what was the pattern running through it?" A teaser up top carries the thread while the month is still unfolding. You read your own lessons before you write the synthesis. Months are the unit the practice actually moves in.

Today’s news

Praxis keeps an eye on the US economic calendar so you don’t have to keep a tab open for it. High-impact releases — the red folder: FOMC, CPI, NFP, PPI — surface where you’ll actually use them: a chip on the Desk that names the next event and opens the full day on hover or tap, dots on the Monthly calendar, and one quiet line on the Dashboard. Alongside the scheduled news, the structural days a futures trader plans around are marked too — OPEX and triple-witching Fridays, market holidays, month- and quarter-ends, and the quiet no-news Mondays. If you want the second tier as well — the orange folder, medium-impact releases — turn it on once in Settings; the red folder always shows. It is awareness, never a signal, and none of it touches a score.

Letter

The Sunday letter: your own reflections from two to eight weeks ago, held up; the identity reading; the week-in-repair paragraph when the week earned one; the numbers against the previous four weeks. Detailed in Reflection & the letter.

Patterns

Everything the evidence engine currently has to say — drift, sustained strength, triggers, risk-vs-1R, repair, and the green-day reads (whether your process slips on profitable days) — each with the record backing it. Dismissed patterns stay quiet for a week. Detailed in Reflection & the letter.

Path

The long view: twenty-six weeks of the two signals that matter — discipline (fast) and identity (slow) — plotted toward the Operating cell where they meet. Weeks with too little data render as gaps, not guesses. The math behind both axes is in How scoring works.

Ladder

Your tier history: every promotion and demotion, the score at the crossing, and the context that was read at the time — clean days, streaks, your strongest and weakest rules at that moment. The ladder is a record of who you have been; the dashboard says who you are.

Milestones

Process prizes, never outcome prizes: ten clean days in a row, a fully clean week, thirty days without a bad loss, fifteen straight sessions with every risk rule kept, the twelve-week Second Nature. They live on the Ladder, grouped under the chapter each one belongs to — so you can see which milestone sits in which tier, and which is within reach. You can define up to five of your own against the same metrics, added to a chapter straight from the ladder; manual ones (the firm funded you, the buffer locked) you mark earned in place. Once earned, a milestone stays earned — it marks that you got there, not that you never left.

Accounts

Your first account is created through a short guided setup — prop-firm or personal (a brokerage, or a sim), stage, starting balance, and the firm's hard floors. Afterward you add and manage every account here.

One active account at a time. Keep as many accounts on record as you trade — firm, stage, and balances — but only one is active, and the active account is the one the daily practice runs against. Activating another archives the rest (their history is kept, not deleted) and makes it your default. The reasoning is the product's: identity and discipline are trader-level — earned once across the practice, not a separate tier per account — so the daily loop reads a single account. Each account carries an optional, structured resolution of its firm floor — a daily-loss-limit and a max-drawdown shape (static or trailing) — so Praxis can show how close you came to today's limit and your distance to the account's floor, without ever pretending to enforce it. Your method — risk per trade, contract cap, session window, and the exit split — no longer lives on the account; it is composed and locked on the Playbook, and that committed playbook is what the Desk’s risk sizer reads.

Risk envelope

Logging two extra numbers on each trade — the worst the open position went against you (maximum adverse excursion) and the best it showed before you closed (maximum favorable excursion) — unlocks a risk-versus-reward read. Praxis expresses it in R, the unit you committed in your playbook: the heat you actually carried against your planned 1R, how much of the favorable move you kept, how much you handed back, and an edge ratio that judges your entries independent of how you exited. None of it touches your discipline score — the risk you take is already scored through your risk rules; these numbers only keep those answers honest. The day's review rolls the per-trade numbers up, and over a month the monthly review turns them into a distribution — led by one plain-English read of what's working and what's leaking: your average heat, how often you stayed within plan, your give-back rate, and your mean edge ratio.

Two whole-day readings can't be derived per trade, so you enter them by hand on the Review: the worst point from the day's open and the max intraday drawdown. These let the page catch a green day that quietly breached your daily loss limit before recovering — the kind of intraday hole a closing figure hides.

The Desk

The one surface that isn’t in the sidebar. Summon it from anywhere with ⌘D (Ctrl-D on Windows) or the small launch tab; it opens as an overlay over whatever page you’re on, and closes with the same key, Escape, or a click outside. It is unavailable on the Review and the trade-logging page — those are the practice itself. The Desk is for the hours in between: the moment of entry.

Everything on the Desk is drawn from your committed playbook. At the top sits your standing identity line (or, if no playbook is committed yet, a prompt to compose one). Below it, Confirm the gates — a checklist of your playbook's per-session critical rules, with a running count and a stamp that holds you to "waiting is the trade" until every gate is clear; your critical rules also show as standing pills. Beside that, a risk sizer: pick the instrument (your committed instruments, or the four micros/minis as a fallback), enter your stop in points and your risk in R, and it returns the contract count, then splits it into the exit tranches from your playbook. It warns you when the size your risk wants would breach your playbook’s contract cap, and shows your account limits (R in dollars, the cap). When the trade is done, a Log this trade button opens the trade form. A today’s-news chip sits in the header.

Nothing on the Desk is scored. The checklist is a gate you pass through, not a thing you are graded on, so its checks reset every time you summon it — a fresh gate each session, never a record. The gates and sizer always reflect your live playbook, including a fresh commit or a pending amendment once it takes effect.

Settings

Who you are — your aim (the declaration the practice serves) and your display name. Practice configuration — your default account, theme, and the email reminders, all sent in your market timezone. Three reminders exist: the Sunday letter note (when a new letter is ready), an evening reminder to log (weekdays, only if you haven’t logged that day), and occasional nudges (on by default — one gentle note if you go quiet for a while, never before you’ve practiced at least once). A separate toggle shows medium-impact (orange folder) US news. The Sunday letter and occasional nudges go out at your chosen morning send hour; the evening reminder uses a separate evening send hour. Away periods — declared date ranges when the practice should not expect you. Away days excuse the ritual-rate denominator, the staleness read, and reminder emails, and nothing else; adding one after the fact is the intended way to repair a rate dented by a trip. Email and Security — address changes and password, each a deliberate two-step. Your data — a one-file JSON export of everything you’ve written. Delete account — permanent, double-confirmed (your password and the word itself), and total: every session, score, pattern, milestone, and shared card goes with it, unrecoverable. Before it takes your password, the page names one thing honestly: wanting to wipe a record after a hard stretch is the same move as refusing a stop.