Frequently asked questions

Composure is a trade journal and performance coaching platform for futures, forex, and equity traders. You log or import trades, and Composure uses AI to score your execution quality, track behavioral patterns, enforce risk guardrails, and generate weekly coaching reports — all without requiring broker login credentials.

Traders who want to build discipline and consistency — from beginners logging their first trades to full-time futures/forex traders. It is built primarily around NQ and ES futures but works for any instrument. It is a journaling and coaching tool, not a brokerage or signal service.

No. Composure works without any broker connection. You can log trades manually, import a CSV or HTML export from your platform, or install the free MT5 Expert Advisor for automatic syncing. None of those three require your broker password. Some direct connections do sign in on your behalf — Tradovate, Rithmic, TradeLocker, dxTrade, Match-Trader and KIS ask for platform credentials, which are encrypted at rest and only ever used to read your trade history.

File import (free, any plan): MetaTrader 4/5, NinjaTrader 8, cTrader, Tradovate, Rithmic, Apex Trader Funding, TradeLocker, dxTrade, MatchTrader, Quantower, TradingView (Paper Trading), Interactive Brokers, Bybit, and Taiwan brokers (永豐/SinoPac, 元大/Yuanta, 富邦/Fubon, 凱基/KGI, 群益/Capital, and more — 台股 and 台指期). Several also support automatic live read-only sync on the Pro plan — MetaTrader 4/5, cTrader, Tradovate, Interactive Brokers, Bybit, dxTrade, MatchTrader, TradeLocker, Rithmic, and 한국투자 KIS (Korea — domestic + US stocks). Full step-by-step guides are at composure.trading/brokers.

Yes. Free accounts include 50 trade entries per month, full journal access, interactive charts, trade replay, the economic calendar, and the full Statistics page — including Ask your trades, which answers questions like “NQ shorts P&L this week”, “which instrument makes me the most” or “long vs short” from your own trades, with no AI and no limit. The Pro plan unlocks the AI layer — trade scoring, Trade Guard, pre-market briefs, weekly coaching reports, broker sync, and an AI follow-up that reads the Ask-your-trades questions the free parser cannot. The figures are computed from your own trades either way — the AI only interprets the question, never the numbers. MT4/MT5 EA sync works on every plan: the first historical import is free regardless of size (or set SyncDays to 0 in the EA to start fresh), and from then on synced trades count toward the monthly entry limit — when you hit it, sync pauses and resumes automatically next month, and nothing is lost. Direct broker connections (SnapTrade, Rithmic) are Pro.

Create a free account, complete the 2-minute onboarding (your instrument, session, and risk preference), then either log your first trade manually or import your broker history. The AI scoring and weekly report activate automatically once you have trades.

Composure is a responsive web app — it works in any modern browser on desktop, tablet, and phone, with no separate download. On mobile you can add it to your home screen for an app-like experience.

Composure supports multiple languages — switch from the globe icon in the sidebar or on the login page. English and Traditional Chinese are the most complete; other languages are still being translated. AI-written text (trade reviews, weekly reports, coach briefs, scenarios) follows your interface language by default; if you want it in a different language, pick one under Settings → AI output language.

Export your trade history from your broker and upload it at Import Trades → 台灣券商 — Composure auto-detects the Chinese column headers, so any Taiwan broker's CSV works. In your broker's website or app open 成交查詢 (executed trades) or 已實現損益 (realized P&L) and export to CSV, then drop it on the upload box. It reads 商品/代號, 買賣別, 成交價, 數量/口數, 成交時間, 手續費, 交易稅 and 損益, and both 台股 and 台指期 are supported. A 成交明細 (fills) export is automatically FIFO-paired into round-trip trades, while a 已實現損益 export is read directly using the broker's own P&L. It is free on every plan, read-only, and never needs your broker password. Supported brokers include 永豐 (SinoPac), 元大 (Yuanta), 富邦 (Fubon), 凱基 (KGI), 群益 (Capital), 中信 (CTBC) and more; automatic API sync (via 永豐 Shioaji) is planned as a next step.

Generate an App Key and App Secret in the KIS Developers portal, then enter them with your account number at Import Trades → 한국투자 KIS — Composure syncs your domestic and overseas (US) closed trades automatically, read-only. It uses KIS's official REST API with an App Key, so your 공동인증서 (certificate) never leaves your own computer and is never uploaded to us. We only read your executed trades to build your journal; we never place, modify, or cancel orders, and the App Key can be revoked in the KIS portal at any time. It is a Pro feature. Support for 키움 Kiwoom is planned as a next step.

Yes, and correcting a trade never loses what your broker said: the original figure stays on record and is shown on the trade's detail page, in smaller text directly under whatever you changed it to. Brokers do occasionally report something wrong and you should not be stuck with it, but you should never have to take your own word for what was there before either. On a synced trade the whole broker record — symbol, direction, quantity, entry, exit, and (only when your broker actually reported them) stop loss and take profit — shows as a receipt with a single "Correct it" button underneath it; press it once and the entire record becomes editable together. There is no separate toggle per field or per leg any more. The edit is never silent: the first time you actually change one of those values, Composure snapshots your broker's original figures and marks the trade as corrected. From then on both numbers are visible — your corrected value as the trade's figure, your broker's underneath it labelled as theirs. The edit form lists every changed field as was-then-now; the read-only detail page shows the pair on each fact that has its own row there, which is all of them except the symbol (that one is the page heading, not a row) and any field you have since left blank. With P/L privacy on, both figures mask together, because the original is money too. A corrected trade stays in your journal and your normal statistics, but stops counting toward your Verified Track Record — evidence someone edited is not evidence. If your broker never sent a stop or target for a trade, that field is a plain, always-editable input from the start — there is nothing to "correct" because there was no broker figure to protect. Entry timeframe and entry reason are never broker facts, so they stay editable at all times regardless of the toggle — no broker reports which chart you were watching or why you took the trade. Notes, ratings and tags are yours to change freely and never mark anything. Broker-synced trades cannot be deleted.

On the trade's edit page. Trade Detail is now a read-only report — your written review and its star rating, your screenshots, your tags, the AI score and coaching note, the automation summary and the market context are all still there to read, plus a "Review state" strip of ticked/unticked chips telling you what the trade is still missing — but nothing on that page is a control any more. Everything you *do* to a trade moved one click away: press "Edit" at the top right of the detail page (on a broker-synced trade that page is titled "Add your thinking"). The form itself is the left column, with "Execution notes" — your written review, the 1–5 star self score and note images — and the market context panel with "Build market context now" underneath it. The review tools are at the bottom of that same left column, below the market context panel, in this order: "Review checklist"; "AI coaching feedback" with the "Get AI score" button; the "Auto-Tag Agent" panel; and "Journal automation status". You reach them by scrolling the page down normally, and the chart stays pinned in the right-hand column the whole way down — so those four are on screen beside the chart, not behind it. (They briefly sat under the chart in the right-hand column instead, where the pinned chart covered them at every scroll position and they could not be seen at all; that is fixed.) On a narrow screen the two columns stack and everything reads top to bottom: chart, then the form, then those four panels. Nothing was removed and no feature changed plan — reading a trade back and changing it are simply no longer the same screen.

Execution discipline, not profit. Hover any star with the mouse, or Tab to it with the keyboard, and a one-line description of that rating appears next to it. 1 star means there was no plan — you decided where the stop went only after you were already in it. 5 stars means you executed exactly to plan, and that holds even if the trade lost money: a trade that followed your process and still lost is a five, because the scale grades whether you followed your process, not whether the trade won. This is the only place a self score lives on a trade — there is no second rating control anywhere else on the edit page.

The detail page groups everything about a trade into four sections, in this order: what your broker reported (times, prices, direction, size, fees), what you planned before the trade (stop, target, timeframe, playbook, entry style, tags), what came out of it (points, P/L, R:R, MAE/MFE), and what you made of it afterwards (your rating, notes, tags, AI score) — so you can tell at a glance which numbers someone else vouches for and which ones you typed yourself. Fields you never filled in are not shown individually; instead they collapse into one line near the bottom of the group, something like "6 more fields not filled in (Playbook, Entry style, Pre-trade plan, …)" — press it to expand every one of them at once. A field is only ever named there if it is genuinely blank: a self-rating of 0, "no" on "did you follow your rules," or an MAE of exactly $0 all count as answers and show normally, not as missing. Fields that do not apply to this trade at all — an exit price on a trade that is still open — are left out of both the visible rows and the missing-field count, since neither answers nor cannot apply are the same as "you forgot." "Which rules did you break?" is no longer listed there at all, because the trade form no longer asks it; what you recorded in it before is still shown on this page. Nothing here is a limit or a paywall; it is purely about not cluttering a report with blanks. Fill any of it in from the trade's Edit page.

A playbook is a decision tree. You build it on the Playbooks page by pressing ＋ next to any step to add the next level, double-clicking a step to rename it, and × to remove it (removed steps are archived, so past trades keep their history). Templates load straight into the tree so you can change them before saving. If you already have your steps written down, "Paste an outline" turns indented text into a tree in one go. When you log a trade with that playbook, the form shows one level at a time as large buttons; each button shows the win rate and trade count for that branch. You can skip a step that isn't marked required and go straight to the level below it — skipping is recorded, not penalised. Open the playbook and pick the "Decision tree" tab to see the tree with data: thicker lines are branches you walk more often, green means that branch wins more than half the time, red less than half, grey means too few trades to say. Hover any step for its numbers; click it to see the trades that went through it, what skipping it changed, and what missing it cost. Trades that come in from a broker have no path yet — add one from the trade's page in three taps. Rules marked with a ⚙ are judged automatically from the trade's own fields — see "Can Composure check some playbook rules automatically?".

Open the trade from your journal. If it has a playbook but no path, the "Path" block shows the tree one level at a time — tap through it and save. The journal list can filter to "No path yet", and the dashboard shows how many trades are still waiting. Paths added after the trade closed still count toward the tree's line thickness and trade counts, but are left out of the "without this step" and "skipping this step" comparisons, so hindsight can't inflate a rule.

Nothing was lost — the field is no longer asked on the trade form, but every answer you already gave is still in your account and still readable in three places: the trade's detail page shows them, the trade list has a "Rule violations" column you can switch on from the "Columns" button, and the CSV export from the Trade Journal ("Export CSV") includes a rules_broken column. The question itself was retired because the playbook rule checklist answers it better and more precisely: instead of picking from a generic list of mistakes after the fact, you tick the specific rules of the playbook you actually traded. New trades will simply leave the old field empty, which is normal and not a fault. The option list for it also stays in Settings, because those labels are what turn your old saved values back into readable words.

Open the playbook and pick the "Decision tree" tab. Each step shows how many trades walked it and its win rate; hover a step for its numbers, click it to see the trades that went through it, the win rate with and without that step ("Without this step"), and, for steps you can skip, the win rate when you skipped it ("Skipping this step"). Steps you skip most often show a higher "Skipped" count on the hover card; steps your trades leave unticked show up as a low walked count. These comparisons only include trades whose path was recorded before the trade closed, so ticking rules after the fact cannot flatter a rule, and they show association, not causation. Branch comparisons are part of the Pro plan; the tree itself and each step's win rate are free. Statistics answers the same question as a ranking, and which card you get depends on the shape of the playbook. A FLAT checklist still gets "Most-missed rules": which rule you leave unticked most often, counted on the same denominator the playbook page uses, so a rule's miss count there and the N/M on a trade agree. A playbook built as a DECISION TREE gets "Branch preference & skips" instead, because "missed" is the wrong question for a tree — once the siblings you did not pick, the levels below where the walk stopped and the steps you skipped are all correctly excluded, a pure single-choice tree can almost never produce a miss, so that card would sit there empty. The tree card answers the two questions that do apply: at each single-choice layer, how often you took each option (with its share of that layer, its win rate and a 95% confidence interval once that branch has enough closed trades), and the steps you skip most often together with the verdict on whether skipping them made a measurable difference. Taking a branch is counted as a fact, so a choice ticked after the exit still counts and the card says how many were; the skip verdicts exclude after-the-fact ticking exactly as the playbook page does. If you have both kinds of playbook you see both cards, each counting only its own books. What counts as a miss is deliberately narrow — a step you skipped on your way to a deeper one is not a miss (skipping is recorded, not penalised), the siblings you did not pick under a single-choice step are not misses (picking one was the point), the levels below where you stopped are not misses because the walk stopped before them, and veto conditions produce no row at all, since correctly not ticking a veto is not a step you forgot. A trade you walked only halfway still counts for the steps it did walk. To get from the ranking to the actual trades, the playbook's Executed trades tab has a Rules column: N/M per trade, "Off-plan" when a veto was ticked or two mutually exclusive options were both ticked, "Unfinished" when the walk stopped above a required step, and — when the trade has no rule data at all. One divergence worth knowing on a flat checklist — a playbook where no step is indented under another — is that this Rules column counts every rule in the book, while the "Rules Followed %" KPI counts only the steps you marked as required, so the same trade can read 3/5 in the column and 100% in the KPI.

Yes — click a playbook's name (or its "Open" button) on the Playbooks page and it opens at its own address, one playbook per page. That page is read-only and gathers everything about that one strategy: the same six KPIs and edge verdict the card shows, its rules — both your free-text notes and the named "Rule checklist", with the "When followed" / "When missed" money beside each rule — and a "Strategy performance" block with a cumulative P&L curve, max drawdown, longest losing streak, R expectancy (with the number of trades that number is based on) and a keep / watch / cut call. Every money figure on that page is net of commissions — the KPI row, the "Strategy performance" block and the per-rule "When followed" / "When missed" columns are all the same convention, so the totals on that page agree with each other and with what the Trade Journal shows for the same trades. Below 5 closed trades the verdict says why it cannot judge yet instead of guessing. There is also a link straight to the Trade Journal filtered to that playbook. Editing has not moved: to rename a playbook, change its description, its free-text rules or its rule checklist, use "Edit" — from the card on the Playbooks page or from the top of the detail page. Sections with no data simply do not appear, so a brand-new playbook shows a short page rather than a wall of dashes.

Commission / fees is an editable field on the trade's edit page, and fixing it costs you the trade's verified status but not your broker's original number — that one stays visible underneath your corrected figure on the trade's detail page. Treat commission as an execution fact rather than a note, because on a broker-synced trade it is one: saving a different figure marks that trade as corrected exactly the way editing a price or a time does, and a corrected trade stops counting toward your Verified Track Record. The form says so right under the field — "Editing this marks the trade as corrected and removes it from your verified track record." The trade itself keeps its place in your journal and your ordinary statistics; only the un-fakeable broker-verified record drops it. On a trade you logged or imported yourself there is nothing to correct against, so no flag is raised — and manual trades never counted toward the Verified Track Record in the first place. Worth getting right either way: commission is subtracted from the net P/L your journal shows, so the figure is not decorative.

Yes — click a column header. Symbol, Entry, Exit, P/L and RR sort; the first click is ascending, the second flips to descending, and the third clears back to newest-first. An arrow marks the column currently doing the sorting. The sort runs in the database across every trade that matches your current filters — not just the rows already loaded — so "sort by P/L" really does surface your single biggest loss even when it sits far past the first page. Columns the list works out at display time (Trade Swing, review status, the latest-note preview, and the session code) deliberately get no sort control: the database cannot order by a number it never stored, and a header that quietly sorted only the rows on screen would answer the question wrong. Setup tags and behaviour tags don't sort either, for a different reason: a trade can carry more than one of each, and there is no honest single order for "breakout-retest + late entry" versus "breakout-retest" alone. Both of them stay filterable from the journal list as usual. Sorting and drag-to-reorder share the same headers without fighting — a press that never moves is still a click.

MAE is the furthest a trade went against you before it closed; MFE is the furthest it went in your favour. Together they show whether you exited near the top of a move or gave it all back — the trade detail page draws this as a bar named "Trade Swing", with MAE and MFE marked at each end, and adds a plain-language line telling you what share of the favourable move you actually captured. Composure gets the numbers from the best source available: NinjaTrader exports carry the broker's own tick-by-tick figures; otherwise you can press "Auto-fill MAE & MFE from chart" on the trade's edit page and Composure computes them from a free historical market-data feed. That auto-fill needs the trade's exit time filled in first, since it defines the window to scan. Every trade says which source was used, because a tick-level figure and a candle estimate are not equally precise — candle data understates extremes, always in the conservative direction. Intraday futures data can be delayed by a few hours or briefly unavailable for a very recent trade; when that happens, or the instrument isn't supported, enter MAE/MFE manually from your own chart instead. Where no price history exists at all for that instrument and period, MAE/MFE are left blank — shown as "—" in the journal list, with the excursion bar simply not appearing on the trade detail page — rather than showing a guessed number.

Composure reads the actual fills behind a synced trade instead of asking you to describe them. Once a trade has three or more executions, an "Executions" section appears on its detail and edit pages (on the edit page it always shows, with an "Edit executions" link so a bad import stays reachable even when there is nothing to show yet). Each row is one execution: quantity, a quantity-weighted "Avg." price (never a plain mean — closing 1 of 4 contracts and then the other 3 is not a 50/50 average), the time, and a running Position column showing exactly how many contracts you held after that fill, so a partial exit reads as "2 → 1 → 0" at a glance. Same-direction fills that print within a few seconds of each other are shown as one execution — a single 4-lot market order that fills in two pieces 200ms apart is one entry, not a scale-in. If you are still holding part of the position, the open runner shows its realized and unrealized P/L together as ONE trade rather than splitting it into two. On the trade form itself, "Add, close, or layer" is worked out automatically from these fills the moment fills exist, so it stops being something you pick; a manually-logged trade with no fills keeps the old picker as a fallback.

Composure needs to match your trade's symbol to a source of price history, and it now resolves the formats brokers actually send automatically: dated futures contracts (e.g. ESU6), continuous contracts (NQ1!), exchange-prefixed symbols (CME_MINI:NQ1!), and fullwidth-character symbols all resolve on their own, so MAE/MFE auto-fill and the Running P&L curve work without you doing anything for most synced trades. If a symbol still can't be matched, open that trade's Edit page, find "Price data symbol" next to the MAE/MFE auto-fill button, and enter whichever symbol our data source recognizes (e.g. "ES"). This only tells Composure where to look up price history for the chart and the MAE/MFE fill — it never changes your trade's own recorded symbol.

Use the "Log Trade" button in the top nav, or Session > Add Trade. Quick mode (a few fields) is the default for beginners; Full mode captures entry/exit, stop, target, R:R, tags, and notes. Connected brokers import trades automatically.

Once the import saves, Composure immediately analyzes your full history and shows your first personalized insights — things like your worst trading day of the week, whether your edge decays after trade #2, or a confirmed positive expectancy. From there you can jump straight to your Trading DNA, full statistics, or set up Trade Guard. You need at least 5 closed trades for insights to appear.

Yes — re-imports are safe. Duplicate trades (matching instrument, direction, entry time, and price) are automatically skipped. You'll only ever see each trade once in your journal.

Open Account > Broker Accounts. Each connected account shows its status, when it last synced, and the broker's own error message if the last attempt failed. Press Sync now on that row to try again immediately — you do not have to disconnect and reconnect, and re-running a sync never duplicates trades. Connected brokers also pull automatically every 15 minutes; MetaTrader accounts push their own trades through the EA instead, so they have no Sync now button.

Your dxTrade domain, a colon, then your login — for example default:john.smith. Most brokers use the domain "default"; if yours issued you a different one it is in the welcome email or on the broker's dxTrade login page. Getting the domain wrong is the usual cause of a dxTrade connection being rejected even though the password is right.

Log trades manually using the New Trade form — it supports any instrument, any broker, and any time zone. You can also use the Composure CSV template. If you need a specific platform integration, email support@composure.trading.

Yes. Composure is built primarily for NQ and ES futures traders but fully supports forex pairs, crypto, commodity futures, and equity CFDs. Use any symbol name your platform uses — Composure normalises it.

Yes. Each trade entry has a notes field, post-trade review section, and screenshot upload. Screenshots are stored securely and linked to the trade permanently.

Click any screenshot thumbnail — in the edit page's screenshot gallery, in the "Screenshots" card on the read-only trade detail page, or the enlarged image shown in the trade media viewport on either page (there, only while it is at its original size — double-click or press 0 to reset a zoomed screenshot first) — and it opens in a full-size viewer. Press Escape or the close button to leave it; you land back exactly where you were on the page, not scrolled to the top. If the trade has more than one screenshot, the left/right arrow keys step through them, wrapping from the last back to the first, with a position indicator ("2 / 5") so you know how many there are. If a screenshot's secure link has expired, the viewer says so instead of showing a broken image — reload the page to get a fresh one.

Yes. On both the trade's detail page and its edit page, whenever a trade has at least one screenshot, a row of tabs — labelled "Trade media" — appears directly above the chart: the first tab is "Chart", and one more tab per screenshot follows it. Click a screenshot tab and the viewport swaps to that image; click "Chart" and it swaps back. The chart is never rebuilt when you do this — its zoom level, drawings and any indicator panes you added stay exactly as you left them, no matter how many times you switch away and back. The screenshot itself zooms and pans in place, so you can read a price level off it while you fill in the form beside it: scroll the mouse wheel over it to magnify toward wherever the cursor is pointing (up to 8×), hold the left mouse button to drag the image around, and double-click to go back to the original size. With the keyboard, focus the image and use plus and minus to zoom, the arrow keys to move it, and 0 to reset. A badge in the corner shows the current magnification and puts it back when you click it. Switching to another tab and back always returns the screenshot to its original size. Clicking the enlarged screenshot itself (not its tab) opens the same full-size viewer used elsewhere — while it is zoomed in, a click does nothing so it cannot open on top of the image you are dragging, so double-click or press 0 first if you want the full-size viewer. With the keyboard, focus the tab row and use the left/right arrow keys to move between chart and screenshots (wrapping from the last back to the first) or Home/End to jump straight to the first or last tab. A trade with no screenshots shows only the chart, with no tab row at all. This is available on every plan — it is not a paid feature.

Download ComposureSync.mq5 from Settings → Security. Install it in your MT5 Experts folder, paste your API Key and Webhook URL, and enable WebRequests for the Composure domain. Every closed trade is pushed to your journal within 60 seconds — no credentials required. ComposureSync is read-only: it never opens, modifies or closes a position. Keep it current — the EA cannot update itself. Builds before 1.27 saved only the first half of a position you closed in two parts, and builds before 1.26 missed any trade held longer than a single 60-second check. Download it again from the same page and re-attach it to a chart: the re-attach backfills whatever the old build missed.

For file import: make sure it's the export format from the guide at composure.trading/brokers. For the MT5 EA: confirm the EA is attached to a chart, WebRequests is enabled for the Composure domain, and your API Key + Webhook URL are pasted correctly. Auto broker sync requires the Pro plan. Still stuck? Email support@composure.trading.

Every New Trade and Edit Trade form has a "How you traded it" section with five optional fields: your multi-timeframe trend read (bull/bear/range for each timeframe you track — a new account starts with Daily, H1, H4 and 15min, and the list is yours to edit under Settings → Templates → "Your option lists", so add H1 there if you don't see it), entry style (e.g. immediate, retest, late, FOMO), market condition (trend continuation, reversal, range, counter-trend), whether you added, closed part, or layered into the position, and rule adherence. That last one is a "Did you follow your rules?" yes/no — unless the playbook you picked defines named rules, in which case it becomes that playbook's tick-list instead and the yes/no is worked out from the ticks. (The old follow-up "Which rules did you break?" is no longer asked; what you already recorded in it is still on the trade detail page, in the "Rule violations" column and in the CSV export.) None of them are required. Four of these five feed real breakdowns, so it's worth filling them in: entry style, market condition, and your higher-timeframe trend read each get their own breakdown under Playbooks → Performance, and rule adherence gets a "Rules Followed vs. Performance" breakdown on the Statistics page — so you can see in your own numbers whether trading against your own trend, or breaking your rules, actually costs you. (Add, close, or layer doesn't currently feed a breakdown anywhere.) A trade you never answer "did you follow your rules?" on is excluded from that stat entirely — left unanswered, never counted as a rule break.

Yes. Go to Settings → Templates → "Your option lists" to add, relabel, and reorder the choices for entry style, market condition, add/close/layer, rule violations, and the timeframes you track for trend. The "rule violations" list is now history-only — the trade form stopped asking "Which rules did you break?" when playbook rule checklists arrived, so that list exists to keep your past answers readable rather than to fill in new ones. New accounts start with a futures/ICT-style preset (immediate, retest, late entry, FOMO; trend continuation, reversal, range, counter-trend; and so on) — edit it down to your own vocabulary any time. Removing an option archives it rather than deleting it: any past trade that already recorded it keeps showing that value, and you can restore an archived option from the same screen later. This is a separate system from the custom fields further down the same Templates tab — those are extra fields you invent yourself and don't feed any breakdown either. Of these five, entry style, market condition, and trend power real breakdowns on Playbooks → Performance (Pro plan and above); add/close/layer and rule violations don't currently feed one anywhere.

Yes, for four of the five process-field groups on the trade form: Entry style, Market condition, Add/close/layer, and Which rules did you break each end with a "+ Add your own" button. Type a new option there and it is created and selected immediately, with no trip to Settings needed; if what you typed already matches an existing option (even in different case), Composure selects the existing one instead of creating a duplicate. Anything added this way also shows up in Settings → Templates → "Your option lists," so you can still relabel, reorder, or archive it later. Trend is the one exception: its per-timeframe chips have no "+ Add your own" button on the trade form, because a timeframe is a fixed time scale rather than personal wording — add a new one from Settings → Templates → "Your option lists" instead, and it appears on the trade form from then on. A timeframe you already added, or already used on a past trade, keeps showing there either way, including one you later archive or delete elsewhere. This does not extend to behaviour tags: the "What you did right" / "Rules you broke" tags used in AI scoring are a fixed list with no inline-creation button — they are a different field from the five process-field groups above, and a custom word typed there would not be counted as either a strength or a mistake.

The Emotion field on a TRADE was retired — not on the trade form, not on the trade detail page, not as a column in the journal list, and not as a breakdown in the Statistics trade tables. (Two controls that look similar are still there and are staying: the "Emotion state" dropdown on a Statistics calendar day, and the mood chips on the dashboard session capture. Both write to your daily journal note, which records how the DAY felt, not how one trade felt.) The reason is that it duplicated the behaviour tags: "Revenge", "FOMO" and "Tilted" under Rules you broke are themselves emotional states, greed overlapped with FOMO, and fear overlapped with cutting a winner early — so the same trade could be counted twice under two lookalike lists with nothing on screen explaining why. Record how you were feeling with the behaviour tags and the post-trade notes instead; both feed AI scoring, Auto-Tag and the weekly report. Nothing you already recorded was deleted: emotions saved on past trades are still stored, and they still come out in the CSV export (Journal → Export) under the "emotion" column, so you keep every word you wrote — you just cannot add or change one from now on.

A playbook can declare the tags it produces (for example, an IFVG-into-CISD-into-FVG setup declares all three), and selecting that playbook on a trade prefills those tags as editable chips — remove any that did not actually happen this time. A playbook only ever REFERENCES tags that already exist somewhere in your journal; it can never mint a brand-new one straight onto a trade. If the playbook declares a tag that is not yet attached to any trade of yours, it is not applied automatically — instead it appears as a one-tap "+ tagname" button beneath the tag field, and only tapping it adds it, which is also the moment that tag starts existing for you. There is no tag management screen: a tag exists for as long as at least one of your trades (or, for suggesting it to you, a playbook) still carries it, and disappears on its own once nothing does.

You do. Tags on a trade are free-form text, so Composure cannot tell a setup name apart from a note like "no-sl" or "us500-long". Open Playbooks → Performance (Pro plan and above) and tick the tags that are real setups — only those earn a keep / watch / cut verdict in Setup Edge Verdicts. Playbooks always count, with no ticking needed. Nothing is ticked by default, so until you choose, verdicts are based on your playbooks alone. Unticking a tag is not a display filter: it deletes that tag's whole verdict history from Setup Edge Verdicts — not just greys it out — until you tick it again, so if a panel you were reading disappears, that's why. Your ticks also show up on the trade form: in the Setup tags field, tags you have declared as setups are listed first and carry a "Setup" badge. That is a label and a sort order only — nothing is hidden, so every tag you have ever used is still suggested, still pickable and still typeable.

It's a chart under the "Running P&L" heading on a closed trade's detail page showing how that trade's P&L actually moved while it was open, not just where it started and ended. The curve is cumulative — realised plus unrealised — so scaling out of part of a position never reads as a sudden loss the moment you locked in profit; it's plotted in gross terms, with a "Gross · Fees · Net" readout beside it for the number that actually hit your account. Composure never draws a shape it can't back with real price data, so what you see depends on how much price history is available for that exact trade window: enough bars gives a full curve with a shaded band showing the range each bar actually moved through; a short trade with only coarse bars available gets an approximate, blocky line, labelled as such; too little history (a bar or two) draws only two unconnected dots at entry and exit — never a line joining them, since a line would claim the trade ran without drawdown that nobody can actually verify; and if no price history exists at all for that instrument and period, no chart is drawn.

Yes — on the desktop trade list, click the "Columns" button in the narrow rail fixed to the table's right edge to open the picker. That rail sits outside the table's horizontally-scrolling area, so it stays visible and clickable no matter how far right you scroll the table — it never gets covered by a column that scrolls underneath it. It offers 40 columns, grouped into Trade basics, P&L, Process, Tags & scores, Source, and Review, with a search box for finding one fast; tick or untick any of them, or use the All / None / Default shortcuts. Nothing changes until you press Apply, so Cancel really does discard everything you just tried. To reorder, drag the real column header left or right — the label follows your cursor, the column being pushed aside highlights, and the rows underneath rearrange live as a preview. Clicking a header still sorts, not drags: a drag only starts once you actually move the pointer. Without a mouse, focus a column header and press Alt with the left or right arrow key. Your layout is saved to your account and follows you back next time. The Symbol column and each row's link to open the full trade report always stay on the table regardless of what you hide, so clicking "None" can never leave you with a blank list. Ten columns are shown by default; the trade list on phones is a card list with a fixed set of fields, so the picker is desktop-only.

Yes — go to Settings → P&L display and change Unit from "Dollars ($)" to "R-multiples." Individual trade rows in your journal and stats then show as a multiple of planned risk instead of a dollar amount, so a 1-lot trade and a 10-lot trade compare on the same scale. Totals and averages always stay in dollars even in R-multiple mode — only individual trade rows switch units, so a dollar total sitting next to an R-multiple row is expected, not a bug. A trade with no recorded stop has no defined R — there is nothing to divide the result by — so it shows "—R" rather than "0R"; 0R would wrongly claim the trade broke exactly even in risk terms.

Privacy (Settings → P&L display) masks P&L numbers in your journal and stats — dollar figures, R-ratios, MAE/MFE, points/ticks, and chart tooltips all render as a fixed placeholder instead of the real value, which is useful while recording your screen. It does not mask entry/exit/stop/target prices, chart shapes, axis labels, which trade, day, or tag ranks best or worst, or your win/loss classification — win rate, streaks, and win/loss counts stay visible either way. Privacy is independent of the Unit setting, so you can combine them, e.g. R-multiples with the dollar amounts masked.

Yes — on a trade's detail page, press the "Share" button (visible only to you, the trade's owner) to open a privacy checklist and generate a shareable link. The checklist lets you choose what to hide — Hide P/L, Hide size, Hide prices, Hide notes, Hide broker — and Hide P/L and Hide size are checked by default, so your profit/loss and position size stay private unless you turn them off. Saving produces an unlisted link at /t/<token>: anyone with the link can view a read-only version of the trade with your hidden fields masked, but the link itself is never publicly listed or searchable — no one finds it without you sharing it. Revoke the link any time from the same panel; once revoked, the public page 404s (not found) for anyone who still has the old link. This is available on every plan and shares one trade at a time — there is no whole-journal sharing. To share a whole playbook instead, use the Share button on the playbook's own page.

Yes. On a closed trade with no stop, the detail page shows a red "Not set — discipline penalty" row, because trading without a stop is the worst habit a journal can catch and it is shown loudly rather than buried in the "not filled in" list. A missing take-profit target is a softer amber note. The test is presence, not timing — a stop you set a few minutes after entry still counts, so only a trade that carried no stop for its whole life is penalised, not one where you set the stop slightly late.

Yes, and both are derived from your trade's real executions — nothing is hand-entered. A scaled exit shows a take-profit ladder on the detail page: each partial close becomes TP1, TP2, TP3 and so on, with the share of the position it banked, its price and its R. If your stop moved while the trade was open, a stop-loss journey shows STOP 1 → STOP 2 → …, marking a move toward entry or to break-even as good and a move away from entry — widening your risk — as a red discipline penalty. Tracking stop moves needs a live feed that reports the current stop, such as the MT5 Expert Advisor, TradeLocker or eToro; trades logged before this existed simply have no history.

It plots your unrealized P&L across the life of one trade, with entry = breakeven as the zero line, so you can see how deep it dug against you, how far it ran in your favour, how many times it crossed back through breakeven, and where you finally exited. It uses the richest data available for that trade: a live sampled price path when there is one (real crossings), otherwise a reconstruction from price bars, otherwise your two MAE/MFE extremes — the chart labels which. Hover any point to read the R and the dollar amount at that moment, and how far the drawdown is toward your stop. It replaces the old separate "Running P&L" chart, which was another view of the same thing.

They turn your MAE/MFE and trade paths into coaching on how you enter and exit, and they also appear on each playbook. Exit Quality shows your capture rate — how much of the average favourable move you keep versus give back — plus a suggested breakeven trigger drawn from how far your losing trades actually ran before failing. The Breakeven Simulator replays the rule "move the stop to breakeven at +X R" over your own trades and shows, per trigger, how many losers it would save against how many winners it would scratch, and the net change in expectancy — so you can see the real cost of going to breakeven too early. Entry Timing Edge estimates whether waiting for a pullback before entering would have paid, and the Typical Path chart shows the median journey of your winners versus your losers, so you can see that winners routinely dip against you before they run. These panels need trades that carry a stop (for the R scale) and are hindsight reviews of your own history, not predictions.

Because R is risk, and risk needs the stop your trade OPENED with. Your platform reports the stop as it stood when the position closed — so on any trade whose stop you moved to break-even, that number sits on the profit side of your entry and is locked profit, not risk. Composure will not divide by it: measured across one real account, doing so produced R multiples as high as 1:377 and an average R of 1:1000, and a wrong number you cannot question is worse than a blank. Three things fix it. The MT4/MT5 EA from 1.34 records the stop at the moment a position opens, including the fact that it opened without one — so new trades carry their real risk. The journal always measures from the widest stop on the risk side rather than the last one reported. And for trades already in your journal with no usable stop, the edit page shows a field called "Stop you were working to": type the stop you actually had and the R appears, labelled "R multiple · your stop" so you can always tell it apart from one measured from your broker's own record. That value is stored separately from what your broker sent and never counts toward your Verified Track Record — the whole claim of that record is that its numbers came from your broker, not from anyone's memory. On a staged or scaled position, R is measured against the most risk the position ever carried at one time, not the first clip's: a position entered in twenty-one parts should read the same as the same risk entered once, and the panel says how many legs it could measure when some had no usable stop.

Two ways: start from a template, or build it yourself. Templates live under "Start from a template" on the Playbooks page — 38 built-in playbooks in nine categories (ICT / SMC, Order flow, Zones, Breakout, Mean reversion, Trend, Options flow, Options, Event). Pick a category chip to see just that group, and tick tag chips to narrow further; ticking two tags shows only the templates carrying both, not either. Every card expands into a "Full strategy spec" — setup, market context, entry trigger, invalidation, stop and target logic, anti-patterns — and in Traditional or Simplified Chinese that description and spec are written in Chinese, not machine-translated. Pressing "Use template" loads the template onto the canvas as two trees, ENTRY and EXIT, with its required steps and veto conditions already marked and its tags pre-filled in the derived-tags field. Step labels stay in English; the explanatory note under each step follows the language you created it in (Chinese for zh-TW and zh-CN, English otherwise). Nothing is stored until you press Save, so edit it first — it becomes an ordinary playbook of yours, not a link back to the template. To build one from scratch instead: press ＋ Starting step and name your first step — or press "Paste an outline", drop in an indented list you already wrote, and Composure builds the whole tree at once. After that you work on the canvas: the + on a step adds the next step under it, Enter adds another step beside the selected one, Tab nests a new step one level deeper, Delete removes the selected step (removed steps are archived, so past trades keep their history), and F2 or a double-click renames it. The arrow keys move the selection from step to step, and Undo / Redo buttons sit above the canvas — ⌘Z and ⇧⌘Z on a Mac, Ctrl+Z and Ctrl+Y on Windows — so trying a structure out is never risky. Give the playbook a name and press Save once; there is no separate rule editor to open afterward, and it always takes exactly one Save, never two.

Yes — on a playbook's own page, press the "Share" button to open a panel and generate a link at /pb/<token>. Anyone with that link, signed in or not, sees the playbook's name, description, derived tags, and the full entry/exit decision tree exactly as you built it — every step, veto, required and single-choice branch, with their notes — but never your trades, your statistics, or your journal notes; those stay out of the public page entirely. A signed-in visitor who presses "Import" gets their own copy of the playbook under the same name, added to their own Playbooks list. It is a snapshot taken at the moment of import: if you edit the playbook afterward, the copy does not update, and there is no link back to your original. Importing the same link a second time does not create a duplicate — it reopens the copy the visitor already has. Turning sharing off from the same panel kills the link immediately, so anyone who still has it gets a 404; turning it back on mints a brand-new link rather than reviving the old one. This is available on every plan. If you don't have an account yet, sign up from the link — after your first-time setup you'll land back on the shared playbook and the import finishes on its own.

Enter adds another step at the same level as the one you have selected; Tab nests a new step under it. Those two build most of a tree, and the rest of the canvas keyboard is just as small: Delete removes the selected step and everything under it (archived, not erased, so past trades keep their history), F2 renames it — a double-click does the same — and the arrow keys move the selection between steps. ⌘Z and ⇧⌘Z undo and redo (Ctrl+Z and Ctrl+Y on Windows), the same two actions as the Undo and Redo buttons above the canvas. If you would rather not build it step by step, "Paste an outline" takes an already-indented list from a spreadsheet or notes app and turns the indentation that is already there into the tree.

No. A playbook you wrote before this changed opens on the canvas with its steps already laid out, and the free-text rules you typed are kept — they sit in a collapsed "Rules (text)" line on the Decision tree tab, not deleted. Nothing is converted in the background ahead of time, and until you open and save a playbook it keeps working exactly as it did before. The first time you press Save, those lines become real, individually-configurable steps (with the parent/child relationships your indentation implied) without losing your original text. Templates are a separate question: a playbook you once created from a template is a normal playbook of yours and is never touched when the template lineup changes. Three templates did change in the current lineup — Momentum Scalp was removed, so an old ?template=momentum-scalp link now opens a blank create form rather than an error; Fibonacci Retracement is now called Optimal Trade Entry (Fib); and FVG Fill is now FVG Fill + IFVG. Every other template id is unchanged, so your saved ?template= links still work.

Thirty-eight, in nine categories: ICT / SMC (8), Order flow (10), Zones (5), Breakout (2), Mean reversion (1), Trend (3), Options flow (2), Options (4) and Event (3). Each card carries a capability badge, and the badge is computed from the template's own description and rules rather than set by hand, so it can never disagree with what the Pine export actually accepts. "Pine-ready" means the conditions are mechanical enough to become a TradingView strategy; "Spec only" means the template leans on reads a script cannot make and stays a journaling spec. Nine are Pine-ready: Opening Range Breakout (ORB), VWAP Mean Reversion, Trend Pullback (EMA), Initial Balance Break & Extension, Composite HVN / Naked POC, VWAP Trend Hold, Anchored VWAP Pullback, VWAP Reclaim (Open Failure) and Weekly VWAP Magnet. The other 29 are spec only. Even a Pine-ready card carries a short line under the badge naming the part that stays discretionary — on ORB the RVOL threshold and the NR7 read are mechanical but judging a clean break is not, and on Weekly VWAP Magnet the macro veto needs you to look at the calendar yourself. A playbook you create from a template shows the same badge the card showed.

Partly, and the line runs between reviewing history and reading a live tape. On NQ, ES and anything else with candle history, Chart Studio gives you volume profile (POC, value area, naked POC), TPO / Market Profile (initial balance, single prints, poor highs and lows), VWAP with sigma bands, effort-vs-result, relative volume and an estimated CVD — all derived from OHLCV candles, which is exactly why they still work on a trade you took last month. What they are not is a true footprint, a DOM ladder, iceberg detection or a liquidity heatmap: those need per-tick aggressor side and live order-book depth, and the live layer that feeds them in Composure is crypto only (Hyperliquid) today — there is no such feed for index futures here. So the ten Order flow templates ask you to confirm absorption, a stacked imbalance or a delta flip in Jigsaw, ATAS, Sierra Chart or Bookmap and then record that read in Composure, rather than implying the number is already on your screen. It is also why most of them are "Spec only": TradingView has no per-tick bid/ask attribution either, so the same conditions cannot be exported as a strategy.

Yes, with one structural limit worth knowing first. The journal is one trade per row: log the option under its symbol, tag it `options`, and statistics, tag breakdowns and playbook paths all work on it the same way they do on a futures trade. What does not exist is any multi-leg or position-group field — a vertical spread is a single row carrying your own net numbers, and the wheel is three separate rows (the cash-secured put, the assignment, the covered call) tied together only by a tag you pick. There is no assignment tracking, no per-leg Greeks and no position grouping. The dealer-positioning side lives on the Gamma page: regime, gamma flip, call and put walls, the magnet strike, expected-move bands, the squeeze-conditions board, vanna and charm, and a strike-by-expiry matrix. It does not have IV rank, VIX term structure or tick-by-tick options flow; its quotes are exchange-delayed by roughly 15 minutes and open interest updates once a day before the open. So where a template calls for an IV-rank read — 45 DTE Premium Selling, for instance — take it from your broker and record it in the trade.

The three Event templates — Data Release Reaction, FOMC Day and Close Auction Flow — read the economic calendar that is already in Composure: an event's impact rating, its forecast, and once published its actual, so the surprise is actual minus forecast rather than a number you remember. Two honest caveats. The forecast is a consensus of economists, which is a proxy for what the crowd expected and not a measurement of how the market was positioned — a release can move hard on a small surprise, and that is the market, not a fault in the template. And MOC imbalance, the order imbalance published before the close that Close Auction Flow is really about, has no feed in Composure: that template asks you to type the thesis yourself and keeps its mechanical trigger to price and relative volume, and its card says outright that its evidence is the weakest of the three.

The canvas at full width with your trade data on it — that tab is for reading the strategy, not editing it (the "Edit" button is still where you change the structure). Above the tree, a period selector and, once you have more than one account, an account filter decide which trades the win rates are drawn from; when a period has no trades it offers all time instead of showing you an empty tree. Veto steps are drawn differently from the rest — an amber bar down the side of the step and a small "Veto" tag, and no win rate, because a veto you overrode is a warning you ignored rather than a step that can be scored. Your free-text rules sit underneath in a collapsed "Rules (text) · N lines" line so they never push the tree off the screen, and on a phone the tab swaps the canvas for an indented list with the same colours and numbers. Back on the Playbooks page, each card shows a one-line path summary — your starting step in bold, then the steps after it — instead of a miniature tree, so you can tell at a glance what the playbook actually is.

Yes — on a step's settings panel (Decision tree editor → select a step → Automatic check) you can hand it to the machine instead of ticking it yourself, for six kinds of rule: entering only inside the session windows you pick (New York AM, London, …) or custom New York-time ranges you type in yourself (say, 09:30–10:00 ET); a maximum number of contracts per trade; a maximum number of trades per New York day; a planned R:R of at least some number, worked out from your entry, initial stop and target; an initial stop no farther than some number of points from entry; and stopping once the day's realised loss has reached a dollar amount. Composure judges those rules from the trade's own fields every time a trade is saved — the manual form, an edit, a path backfill, or accepting a playbook suggestion on a broker-synced trade — so there is nothing extra to do. A rule that passes is ticked for you; one that fails counts as broken (on a veto step, "passed" means the veto condition fired, shown as "Veto triggered" rather than "Followed"); and one Composure cannot judge — no stop recorded, say — is simply left out of that trade's denominator rather than counted as missed. Automatic rules carry a ⚙ badge and cannot be ticked by hand in the trade form or the intraday checklist, nor can you override what they decided. A broker-synced trade you never walked through a playbook still gets judged once you attach one — its overall "did you follow your rules?" stays blank, since only the machine looked at it and you did not, but the per-rule stats still count it. Changing a rule's automatic setting only affects trades saved from then on; an existing trade is re-judged the next time you edit it. Ten of the built-in templates ship with auto checks already set on their time-window and R:R steps, and importing or forking a playbook copies its automatic settings along with everything else. It is available on every plan.

It keeps the checklist short by only asking about rules that actually apply to the trade you are logging. Mark a rule "Single choice" and give it sub-rules — say, entry setups iFVG and CISD — and picking iFVG when you log a trade only offers iFVG's own follow-ups (like Breaker Block), not CISD's. That way a playbook can cover dozens of setups without every trade's checklist listing all of them.

A rule marked as a veto is a reason not to take the trade at all — for example, "HTF resistance directly above." Ticking a veto never blocks you from saving the trade; Composure will not stop you from logging what actually happened. What it does do is mark the trade as off-plan, and veto conditions are left out of your "Rules Followed" percentage entirely — a veto you overrode is a warning you chose to ignore, not a step you forgot, so it is not scored the same way a skipped required rule is.

Building and using the playbook tree — writing rules, nesting them, marking one required or a veto, linking a shared step into two branches, colors, notes — is free on every plan, including Free. Branch analytics, the per-step and per-branch win-rate breakdown on a playbook's Performance tab, is a Pro feature: it runs ongoing statistics on top of your tree rather than just storing its shape.

The Composure Score (0–100) is a composite rating of your trading discipline. It weighs execution quality (entry timing, stop placement), rule compliance (Trade Guard adherence), consistency (trading days and patterns), and AI session review scores. A higher score means more disciplined, rule-following trading.

After you save a trade or complete a session review, our AI engine analyses your entry, exit, risk:reward, notes, and recent pattern history. It returns a 1–10 execution quality score and a plain-language coaching note. Behavior tags split into two groups — "What you did right" (calm, decisive, patient, disciplined, confident) and "Rules you broke" (revenge trade, FOMO, moved stop, etc.) — and only the second group can ever lower your score or show up as a costly habit. Tagging yourself as calm or disciplined is never held against you.

Every Sunday, Composure generates a personalised report summarising your week: win rate, average R, Trade Guard compliance, worst behavioral patterns, and a specific recommendation for next week. It's sent by Telegram and also available under Reports. The weekly report is a Pro feature.

Auto-Tag uses AI to suggest setup tags (e.g. breakout-retest, vwap-reclaim) and behaviour tags (e.g. fomo, moved-stop) from your trade notes. You run it from the "Auto-Tag Agent" panel on the trade's edit page — press "Edit" on the trade detail page to get there; the detail page shows the tags you accepted but no longer carries the controls. You review and accept the suggestions — nothing is saved without your confirmation. It requires a Trader plan or higher.

No. All AI outputs are educational summaries of your own historical trade data. They are not investment advice, trading signals, or recommendations to buy or sell any security. Composure is a journaling and coaching tool — all trading decisions are yours.

Scenario AI generates a daily market narrative for NQ/ES based on macro conditions, key levels, and recent news. It describes possible intraday scenarios (breakout, fade, chop) to help you prepare — it is labelled "simulated scenario, not a trading signal" because it cannot predict the market.

Ask it in plain language, in any of the interface languages. It understands a figure, the trades to compute it over, and whether you want them ranked. Figures: net P&L, win rate, number of trades, expectancy, average win, average loss, profit factor, best and worst single trade, average R, average hold time, max drawdown, longest winning and losing streak, commission paid, how much you left on the table after exiting, how often you move your stop, and your average self-score. Narrow it by an instrument, by a market named in words (gold, silver, precious metals, oil, crude, natural gas, indices, stocks, options, futures, CFDs, forex, crypto, bonds, grains), by long or short, by winners/losers/breakeven, by whether you ticked your rules, and by a time window ("this week", "last month", "last 30 days", "today"). Or ask for a ranking: which instrument, which market type, long vs short, which weekday, what time of day, which playbook. A market covers every symbol you trade in it — ask about gold and it answers over XAUUSD and GC together, which no single ticker could. Whatever it understood is shown back to you as chips above the answer, so you can see it read the question the way you meant it. This is free on every plan with no limit and no AI: the figure is computed from your own trades in your browser. When a question is beyond the parser, Pro adds an AI follow-up — and all it is allowed to do is translate your question into the same kind of query, which is then computed the same way. It never produces a number itself.

Trade Guard is Composure's behavioural risk management system. Set a daily loss cap (e.g. $500) and a maximum trades per day. If you hit either limit, Trade Guard activates a two-stage friction system: first a math challenge + 30-second cooldown, then a full session lock with Telegram notification.

Stage 1 (limit reached): A full-screen overlay appears with a math challenge and 30-second timer. You must solve it to proceed. Stage 2 (confirmed override attempt): The session locks for the rest of the trading day. You receive a Telegram notification and cannot dismiss the lock until midnight reset.

Yes. Go to Trade Guard → Settings and remove your daily loss cap and trade limit. The guardrails only activate when you've set limits. You can also configure "observation mode" to track violations without locking.

Trade Guard tracks trades you log in Composure — manually, via import, or synced from a broker or EA — and alerts you by push/Telegram the moment a limit is hit. It does not block orders at your broker. On MT4/MT5 there is one optional exception: if you install the separate ComposureGuard EA and turn on EA auto-close in Trade Guard, it closes positions you open while locked (only ones opened after the lock started, never earlier ones, and it never opens anything). You are notified of every close, and turning the switch off stops it immediately. Without that switch on, no Composure EA touches your positions.

A 30-day summary of what the guard actually did for you: how many tilt trades (revenge re-entries, size escalations) were flagged, the realized dollar cost of those trades, how many warnings fired, and how many days ended with warnings but no breach. Every number comes from your own logged trades and guard events. The single "avoided loss" figure is explicitly labeled an estimate, based on your own average tilt-day loss — never a made-up projection.

Trade Guard is included in Pro — the one paid plan.

A self-serve tool that mirrors trades between YOUR OWN connected accounts — for example from one account to one or more of your other accounts — with risk normalisation (fixed USD, % of balance, multiplier, mirror, fixed lot, or balance ratio).

No. Composure does not manage money for you and there is no managed-account service. Copy trading is a tool you operate between your own accounts; all decisions and risk settings are yours.

Go to Trade Copier, connect a source and destination account, set the risk model and limits, and validate in dry-run first. Going fully live is enabled per account once the setup checks out.

Copy trading is included in the Pro plan.

A broker-verified, un-fakeable trading record. It counts ONLY trades that came from a connected broker sync (real fills) — manually entered or CSV-imported trades never count toward it. Integrity rules: all synced accounts are aggregated (you cannot cherry-pick accounts), it always shows all-time AND the recent 90 days (you cannot cherry-pick a window), results are normalised to R and % (no dollar amounts shown publicly), and a discipline grade (A–F) only appears after 50+ verified trades. The grade is risk-adjusted (system quality + consistency), so a lucky win streak cannot inflate it. View yours under Review > Track Record.

Yes — it is opt-in. Turn on "Public track record" in Settings (or from the Track Record page), then use the Share button to copy your public link (composure.trading/p/…). The public page shows only normalised R / % — never dollar amounts.

Your mentor can build a scoring rubric for you — their own named criteria, like entry timing, risk sizing, or plan adherence — and rate each of your trades 1–5 per criterion (with an optional note) from their own view of that trade. The rubric is per-student: your mentor defines a separate set of criteria for each student they coach. You see your mentor's scores as a read-only "Mentor's assessment" card on your trade's detail page — you cannot edit them. This is a completely separate axis from your own self-score / Composure Score: two different numbers, your mentor's coaching assessment and your own execution self-rating, never combined into one score. If your mentor retires a criterion later, it is archived rather than deleted, so your past scores against it still show. This requires your mentor to be on the Pro plan and an active mentor relationship between you (the existing Mentor feature) — without both, the scorecard simply does not appear.

Yes — press Reply under any comment on a trade to post a threaded reply (one level of nesting). Comments are a mentor feature: only the trade's owner and their active mentor can comment or reply on it. Whenever either side comments or replies, the other party is notified three ways — an in-app alert in the notification bell, a browser push notification, and an email — and this works in both directions (mentor to student and student to mentor). As the student, you can turn off the mentor-comment email under Settings → Notifications → Email notifications → "Mentor trade comments" without losing the bell or push alert; there is currently no email off-switch for the mentor's side.

Your mentor's dashboard for you shows win rate, profit factor, average R multiple, discipline (your self-score), trade count, an equity curve, and your most-missed playbook rules, once you have an active mentor relationship and are both on the Pro plan. You control what shows up there from Settings, in the Mentor section: toggle "Hide P/L" and every dollar figure disappears from your mentor's view of you — net P&L, average win/loss, and the whole equity curve are all masked — while win rate, R-multiple, profit factor, discipline, and your missed rules stay fully visible, so your mentor can still coach your risk management and rule discipline without seeing your account size. You can separately hide position size, entry/exit prices, notes, or broker details too, each its own toggle. This is enforced at the database level, so your mentor can never see something you've chosen to hide.

Yes — from their view of your trade, your mentor can draw marks on the chart with the same drawing tools you have (trend lines, boxes, and the rest) as feedback. You see those marks on your own trade chart in a distinct color labeled "Mentor", and they are read-only: you can't move or edit the mentor's drawn shapes. A "Show mentor's markup" toggle lets you hide or reveal them at any time. Each mark can carry a note from your mentor and support its own discussion thread — you can reply on that specific mark, and replies notify the other side the same way trade comments already do. Either of you can mark a mark "Resolved" once it's addressed (a resolved mark is shown de-emphasized in the thread); resolving is a separate action from replying, so a mark can be discussed any number of times without being resolved, or resolved without a reply. Your own chart drawings are completely unaffected and stay fully editable. This requires your mentor to be on the Pro plan and an active mentor relationship between you (the existing Mentor feature).

One paid plan, kept simple. Free: 50 entries/month, the full journal workflow, interactive charts, trade replay, economic calendar, and full statistics. Pro ($39/month or $348/year): everything — unlimited entries and AI scoring, Auto-Tag, Trade Guard, weekly AI report, Scenario AI, pre-market briefs, automatic broker sync, AI session review, Mentor Mode, and copy trading. During launch, Pro monthly is discounted to $29/month — subscribe while the discount is live and your renewals keep that price. See the Pricing page for details.

The Free plan is the trial — no card, no clock. Journal up to 50 trades a month with full statistics, for as long as you like. When you want the AI layer (trade scoring, Trade Guard, weekly reports, broker sync), upgrade to Pro; there is no separate time-boxed trial to manage.

Yes. Cancel with one click from Settings → Billing. Cancellation takes effect at the end of your current billing period. Your journal data stays on the Free plan permanently after cancellation.

All journal entries, statistics, playbooks, and settings are preserved. Access to paid features (AI scoring, Trade Guard, weekly reports, full stats) pauses. Your data is never deleted when you downgrade.

Yes. Pro is $348/year — that works out to $29/month, saving $120 (about 3 months) versus paying monthly. Switch to annual from Settings → Billing or on the Pricing page.

Payments are processed securely by Paddle, our payment provider. Composure never sees or stores your full card details. Manage your subscription anytime from Settings → Billing.

If you are not satisfied with your first charge, contact support@composure.trading within 14 days for a full refund. After that, refunds are handled case by case — and the Free plan means you can evaluate Composure as long as you like before paying.

Yes. Share your invite link from the Dashboard or Settings. You earn rewards for every paying subscriber you refer. There is no cap on referral rewards.

Yes, on Chrome, Edge, Brave and Firefox 151+. Press "Pop out" on the checklist and it opens in a small window that stays above other applications and remains visible when the browser is minimised, so you can tick items mid-session without switching windows. Two honest limits: Safari does not support this at any version, and on macOS a window will not draw over an app in native fullscreen (use maximised instead of the green fullscreen button). The floating window also closes if you close or reload the tab that opened it — reopening takes one click. It is available on every plan.

Connect Telegram in Settings → Notifications. Once linked, Trade Guard locks, the weekly report, and strategy alerts can be delivered to your Telegram.

Breaking-news alerts are deliberately rare. A headline only reaches your phone if it scores at the very top of the impact scale AND names something that justifies an interruption — a crisis event, a trade-policy action, or a policy-maker like the Fed chair. Everything else still appears on the News page with its impact score; it just does not buzz. If two sources report the same event within half an hour, only the first one alerts. News from Chinese-language sources is shown to you in English.

In Settings → Notifications, enable Web Push and allow notifications when your browser asks. You'll then get real-time Trade Guard and strategy alerts even when the tab isn't focused.

The weekly report needs a minimum number of trades that week and the Pro plan. Also check that Telegram is connected (Settings → Notifications). The report is always available under Reports too. Still missing? Email support@composure.trading.

Use "Forgot password?" on the login page to reset by email. To change it while logged in, go to Settings → Security.

Update your profile in Settings. For email-address changes that don't stick, email support@composure.trading from your current address.

Yes. Use the Export option in the journal/stats area to download your trades. Your data is yours.

Yes. Journal, Statistics, and Export support filtering by account (multi-select, plus "all" and "unassigned"), so you can view each prop/live/demo account separately or combined.

Because one decision is one trade. When the same entry lands on several accounts within a second or two at nearly the same price, Composure treats those fills as a single decision: trade count, win rate, "trade #3 today", streaks and your Trade Guard limits all count it once. The money is never collapsed — P&L, commission and size are added up across the accounts, so your totals still match your broker. The rule that keeps this honest is that a decision can hold at most one fill per account, so scaling into a position on one account stays several trades. Per-account views are unaffected: the account filter, the trade list and prop-firm drawdown always show every individual fill.

First check your plan includes that feature (AI scoring works on every plan — Free includes 15 scores a month; Trade Guard needs Pro). Reload the page. If it still fails, email support@composure.trading with what you clicked and any error you saw.

Yes. Your journal is private by default — only you can see your trades. Row-level security at the database layer means only your authenticated session can read your data. Even the Composure team cannot read your journal entries.

No. We do not sell, share, or monetise your personal data or trade history. The only third-party services that process your data are Supabase (database), Anthropic (AI scoring via their API), and Paddle (payment processing).

Go to Settings → Security → Danger Zone. Submitting the deletion form sends a confirmation email. Your account and data are not deleted immediately — a team member processes the request after you confirm via email. If you change your mind, ignore the email.

Yes, on the Pro plan. Link or add a funded account and Composure evaluates it against its firm's rules in real time — daily loss limit, max drawdown (static, end-of-day trailing, or intraday trailing), profit target, minimum trading days, and the consistency rule — and shows how close the account is to a breach and to a payout. It ships with starting values for common firms (FTMO, Apex, Topstep, MyFundedFutures and more) that you edit to match your exact program, because firm rules change often. These are seeds to verify against your firm, not a guarantee.

Yes. As your funded accounts refresh, Composure watches how much of each daily-loss and drawdown limit you have used and alerts you before you blow the account — a warning when you have used 80% of a limit and a critical alert at 90%. Alerts appear in the notification bell, and as a web-push notification if you have push turned on, so you get a heads-up while there is still room to stop. Breach alerting is part of the Pro prop-firm tools.

Yes. The Broker Accounts page shows a cross-account roll-up across all your funded accounts — which one is closest to blowing up and which is closest to a payout — so you can manage a whole book of prop accounts from one place instead of checking each firm's dashboard separately. Each account also has its own gauges showing distance to breach and progress to payout.

Key Figure Monitoring (the Intel page) tracks influential market figures — central bankers like Jerome Powell, institutional investors like Warren Buffett, corporate executives, and members of Congress. Every figure has an Activity tab: their latest market-relevant news, each with a short AI summary and a bullish/bearish read. Depending on who they are, they also get a data tab drawn from official public filings — investors and 13F-filing companies get Moves (quarterly holdings changes), company executives get Insider (their own-company trades from SEC Form 4), central bankers get Policy (next meeting countdown and stance), and members of Congress get Congress (their disclosed stock trades). You can follow any figure for in-app and push notifications, and request figures you want added.

Yes — it all comes from official, free public sources, and every row links back to the original filing. Investors and companies that file a 13F (e.g. Berkshire, NVIDIA) show real quarterly holdings changes from the SEC's EDGAR database — new buys, adds, trims and exits versus the prior quarter, with an allocation breakdown and an AI summary. Company executives show their Form 4 filings — trades in their own company's stock, where open-market buys and sells carry a signal while grants, option exercises and tax withholding are routine. Central bankers show a countdown to their next scheduled policy meeting (the official Fed / ECB / BoJ / BoE calendar) plus an AI-inferred hawkish/dovish leaning. Members of Congress show their STOCK Act trade disclosures — ticker, buy or sell, and the disclosed dollar range. Important limitations: 13F is a quarter-end snapshot disclosed roughly 45 days late; congressional trades are amount RANGES disclosed up to ~45 days after the trade; and central-bank policy moves all macro assets — equities, gold, the dollar, bonds, FX — not just index futures.

Yes. Open any member’s page from the Congress trades table and press Follow. When that member’s next filing is published you get a notification in the bell at the top right, plus a web push if you have notifications enabled. Alerts only fire for filings from the last 30 days, so loading historical data never floods you. Press Following again to stop. Note what this can and cannot be: members report up to 45 days after they trade, so an alert tells you what was disclosed today, not what was bought today.

Each member has a public page with their disclosed trades and volume, the return on the positions we could match, the S&P 500 over the same days those positions were held, the excess over it, the median holding period, and win rate stated twice — once for closed positions and once for positions still held. The method is on the page because it decides the numbers: filings disclose a dollar RANGE rather than a share count, so each trade is sized at the midpoint of its range and converted to shares at that day’s close; buys are matched to sells oldest-first; a sale with no matching purchase on file (the position predates our record) is counted and excluded rather than scored; only stocks are scored, because an option’s payoff is not the underlying’s price move, so options, bonds and funds are listed with their counts shown; tickers with no price available (delisted, renamed, or mistyped in the filing) are excluded and counted too. Where a member sits on a committee that oversees the industry of a stock they traded, the page flags it — that is a fact about the committee and the sector, not an accusation, and broad committees such as Appropriations and Ways and Means are deliberately not flagged because they touch everything.

Yes — the Congress trades tab in Key Figures (also public at /congress, linked from the website menu) lists every trade that members of the U.S. House disclose under the STOCK Act, parsed directly from the official House Clerk filings and refreshed every few hours. It is free and public, no account needed. You can filter by ticker, member, buy or sell, party, owner (the member, a spouse, joint, or a dependent child), asset type, amount range and trade date, and every filter stays in the link so a filtered view can be shared. It also shows net buying and net selling by ticker over the last 60 days, the most active members over 90 days, and how many days each trade took to be disclosed — anything past the 45-day legal deadline is flagged. Amounts are the disclosed dollar ranges, not exact figures, and totals use the midpoint of each range. Senate trades and handwritten paper filings are not included yet.

The Gamma page (under Live Chart in the sidebar) shows Gamma Exposure — a daily map of how options dealers are positioned and how their hedging is likely to shape the session, across 245 symbols: the index complexes (SPX, NDX, RUT, VIX), the major ETFs, and roughly 180 single stocks, searchable with futures spellings like /NQ or /ES that resolve to the right option complex. It shows a regime badge (positive gamma = dealers dampen moves, range-day behavior; negative gamma = dealers amplify moves, expect expansion; plus a transition band near the flip), the Gamma Flip level, the Call Wall (largest positive-gamma strike, natural resistance) and Put Wall (largest negative-gamma strike, natural support), per-strike bars with wall heat bands over the intraday tape, expected-move bands, today's level history with session replay, and plain-language signal cards. A Screener view flips the chart into a sortable table of every core symbol's regime, flip distance, net GEX and 0DTE share — the morning sweep for which market has structure today. An Expiry matrix toggle opens a strike-by-expiry heatmap with per-expiry GEX and OI totals, and an expiry-combination filter (0D / ≤1W / ≤1M presets or per-date toggles) recomputes the ladder live from just the selected expiries. An IV-simulation slider shifts the whole vol curve up to ±20% and redraws the ladder with dealer gamma recomputed per contract, and a Squeeze-conditions card lights up five structural preconditions for an expansion move (negative gamma, a wall under live test, flow repositioning, 0DTE fuel, one-sided open interest) — deliberately a checklist rather than a probability score. Newer layers: a per-strike IV-skew curve you can toggle over the ladder, a six-state regime strip across the session tape, the 5-day ATM-IV drift, and an expected-move hit-rate readout that fills in as history accrues. Optional overlays: front-expiry 40/25/16Δ delta rails, a Murrey octave grid, and a ladder metric switch (net GEX / DEX / volume / open interest). Levels can be shown on the ES / NQ futures scale with a live basis adjustment, so futures traders never convert by hand. GEX describes market structure and context — it does not generate entries or predict direction.

Each contract contributes gamma × open interest × 100 × spot² × 0.01 (dollars per 1% index move), with calls counted positive and puts negative under the standard dealer-positioning assumption — the same convention used by SqueezeMetrics and other public GEX tools. Snapshots are computed from the full listed option chain roughly every 15 minutes during the US cash session, plus an end-of-day snapshot. The underlying quotes are exchange-delayed by roughly 15 minutes on the current feed, so the displayed spot can trail the live futures price — most visible during fast moves (a real-time feed is on the roadmap); open interest itself only updates once per day before the open, so same-day 0DTE positions are not in it — treat intraday moves in the levels as the market repositioning, not tick-by-tick flow. The 16 most-traded symbols stay pre-computed on that 15-minute schedule; any other symbol computes on first open in a few seconds and is then cached for everyone. Free accounts see the previous session's end-of-day SPX snapshot; Pro unlocks the full symbol universe with intraday refresh and the futures-scale toggle.

Yes — that is the point of it. Every closed index-futures-style trade (NQ, ES and friends) is stamped with the gamma environment it was entered in, looked up from the snapshot that was current at your entry and stored permanently, and shown as a colored chip on the trade's detail page. Statistics gains a “GEX Regime Performance” breakdown — your trades, win rate and P&L in positive gamma, the transition band, and negative gamma — which answers where your strategy actually has an edge. The Pre-Market Plan page shows the last session's structure (flip and walls for NQ and ES), Trade Guard raises a size-down advisory on negative-gamma days, and the AI briefs (Scenario AI, the pre-market voice brief and the per-plan brief) all read the same dealer-structure block as context. After the close, a post-session gamma recap (on the Gamma page, Pro) writes the day's regime story against your own trades, and a full Gamma Exposure course lives in University. Stamping starts from the day GEX launched — older trades have no snapshot to look up, so they are honestly left unmarked.

Composure Cam is an optional tool on the Trade Guard page. With your permission it uses your webcam to read behavioural signals such as blink rate, eye closure, gaze, and brow and jaw tension, and shows a live calm-versus-agitated reading while you trade. It reports behaviour, not emotions, and is only for your own self-review. It stays off until you turn it on, and you can stop it at any time.

No. Composure Cam runs entirely in your browser on your own device. No video, image, or facial data ever leaves your computer - only the derived composure numbers, such as a score and event counts, are saved so they can be compared with your trading. It is not a medical or psychological diagnosis and is never shared with anyone.

Behavioural signals only: blink rate, eye closure, gaze stability, and brow and jaw tension, all computed on your own device. You can optionally add vocal arousal (loudness from the microphone - audio is analysed live and never recorded) and physical-motion arousal, which needs your upper body in frame. Together they produce a live calm-versus-agitated reading, and the Calm indicator feeds the tilt meter on the Trade Guard page so behaviour and risk sit side by side. It reports behaviour, not emotions, and it is not a medical or psychological diagnosis.

No. Composure Cam is built for self-review only and there is no way to share it - no employer view, no prop-firm export, no team dashboard. The video never leaves your device at all, and the derived composure numbers stay in your own private account alongside your journal. This is a deliberate product boundary rather than a setting: workplace emotion inference is restricted under the EU AI Act, and Composure will not build the feature that would make it possible.