ANY SIGNAL.
YOUR RISK RULES.
HONEST SCORES.
You already follow signals from somewhere — a Telegram channel, a friend, your own analysis. SentryQ doesn't ask you to abandon them. It sizes every one from your own stop at your own risk, runs all twenty-six guards over it, hands you a factual synopsis of what you're about to take, and tells you the truth about which sources are actually worth following.
THE CALL WAS FINE.
THE SIZING KILLED YOU.
Two traders take the identical signal. One risks 0.7% and survives the losing run. The other eyeballs "0.5 lots feels right", hits a 90-pip stop on gold, and loses eight times what they planned. The signal was never the variable that mattered most.
Following signals manually
- Position size guessed per trade, so risk is wildly inconsistent
- You take the entries you happen to be awake for
- Nothing stops five signals stacking into one currency
- Provider's claimed record is unverifiable and survivorship-biased
- Fat-finger risk on every manual entry, at speed
Blind auto-copy bots
- Executes everything, including obvious parse errors and recaps
- No sanity check that TP/SL are even on the correct side
- No exposure, news, spread or drawdown guards
- No record of which source produced which outcome
- You find out it misfired after the money's gone
Signals through SentryQ
- Every signal re-sized from its stop at your risk %, in your currency
- Geometry validated — wrong-side stops and recap messages rejected
- All twenty-six guards apply, exactly as they do to strategies
- Confirm-first: you see the exact plan and a factual synopsis before anything is placed
- Every source scored against real prices and ranked publicly
- Multi-target signals split into tranches sharing one stop
PASTE IT. OR
BUILD IT.
Two ways in, on both desktop and mobile. Paste the raw text of a signal exactly as you received it, or build one field by field. Either way the engine does the arithmetic you'd otherwise do badly at midnight.
Real posts do not arrive tidy, so the reader copes with the mess: punctuation in the levels (TP1. 4014, SL. 4031), entry zones attached to the side (Sell 4018/4021), superscript take-profit labels (TP¹ ²) that gold channels love, and providers who write GOLD or SILVER rather than the pair.
It also knows what isn't a signal. A post reading "TP1 Hit, +30 Pips" quotes the same levels as the original call, and a naive reader would open a fresh trade on somebody's victory lap. That one is rejected on sight.
Pasted or forwarded from Telegram, it is the same parser and the same guards. An idea from outside gets no shortcut past your own risk rules — your exposure caps can refuse a trade the sender happily took. And because a refusal is a decision made on your behalf, we measure what those decisions are worth: every entry a guard turns down is replayed against real prices, including the ones that would have won. What our guards actually cost →
PARSE. VALIDATE.
SIZE. CONFIRM.
Parse. The reader handles the messy reality of how signals are actually written — comma thousands, "TP1." punctuation, entry zones, emoji, mixed case, pair names with or without slashes.
Validate. Geometry is checked before anything else: is the stop on the correct side of entry? Are the targets in the right direction? Is every level within a sane distance of the current market? A "TP1 HIT ✅" recap message is recognised as a report, not an order — and rejected.
Size. The stop distance becomes a position size at your risk percentage, converted through the pair's pip value into your account currency. Multi-target signals are split into tranches that share one stop, so partial exits work properly.
Confirm. You get the full plan — lots per tranche, cash at risk, which guards passed — and nothing is placed until you approve it. Every path through this system is confirm-first.
A SECOND OPINION
ON EVERY SIGNAL.
Someone hands you a trade. Is the stop realistic for that pair right now? Is it fighting the trend? Is there news in forty minutes? Does it double a bet you already hold? SentryQ answers all of that before you decide — in plain English, on the same screen you approve from.
Note what it does not do: pick a side. The same restraint runs through the product — ask the desk about a chart and it is expressly forbidden from giving a verdict on whether to trade, and a shared trade carries its numbers with the note that they are numbers, not advice. Why we built it to refuse →
Market context
Risk:reward per target, the stop as a multiple of the pair's own hourly range, trend alignment, distance from the current price, spread as a share of your stop, and the next high-impact event on the calendar.
The setup it reads as
Pullback continuation, breakout, fade of a stretched move, range trade — inferred from where the entry sits in the recent range, how far price is from its average, and whether volatility is expanding. Plus what that shape depends on to work.
Source track record
For signals from a followed channel: that source's measured results overall and on this specific pair — win rate, average R, profit factor — with the live-versus-published-history split shown honestly.
Your portfolio
What the trade would do to your currency exposure against your cap, and whether it adds to or offsets positions the engine is already running on that pair.
Every line is something measurable: a ratio, a multiple of ATR, a distance, a scheduled event, a past outcome. Where a fact has a mechanical consequence, the card says what it is — "a stop inside normal noise can be hit by ordinary movement rather than by the idea being wrong." What you will never find is a score, a rating, a probability, or a recommendation. SentryQ is software, not an adviser, and the decision to take a trade is always yours.
A single number out of ten hides everything that matters. "Stop is 0.47× the pair's hourly range" tells you what to check; an 4/10 tells you nothing you can act on, and invites you to outsource a judgement that should stay yours. The synopsis is deliberately built to make you a better judge of a signal — not to replace your judgement with ours.
THE CHANNELS YOU
ALREADY FOLLOW.
Forward to the bot
Forward any signal message to the SentryQ bot. Seconds later you get a sized preview back in Telegram — lots, cash at risk, guard verdict — with Place and Dismiss buttons. Nothing executes until you tap.
Follow a channel
Search the Sources directory, add a public channel, and SentryQ joins it, imports its recent history, scores it, and subscribes you. New signals then arrive in your app automatically.
Sized on arrival
Every incoming signal is parsed, validated and priced against your account and your risk settings before you ever see it — so what you're approving is a real plan, not a screenshot.
Alerts your way
New signals and their outcomes reach you by push notification, email or Telegram — whichever channels you enable — alongside your engine's own trade and guard events.
SentryQ works with public channels and messages you forward. We do not rebroadcast paid private groups, and we don't ask you to hand over your Telegram account to scrape them. Redistributing a provider's paid content isn't ours to sell — any premium sources we add in future will come through licensed partnerships with the providers themselves.
HOW WE SCORE
A SOURCE.
This is the part nobody else does, so it's worth being precise about. Every claim on our leaderboard is reproducible from the same public data anyone else could gather — we just gather it at the moment of posting, which is the only moment that can't be gamed.
| Step | What we do | Why it matters |
|---|---|---|
| 1. Record | Each signal is captured and timestamped the moment it is posted, before any outcome is known | History can't be rewritten. A provider can't delete the losers later — we already have them. |
| 2. Validate | Geometry sanity checks; ambiguous or malformed messages are marked unresolvable rather than guessed | A mis-parsed decimal can invent a fake +400R winner. We'd rather score nothing than score fiction. |
| 3. Replay | The signal is walked forward against real market bars — fills, stops and each take-profit in sequence | No "would have" estimates. Either price traded there or it didn't. |
| 4. Normalise to 1R | Every outcome is expressed in units of risk, not pips or currency | Makes a 12-pip EURUSD stop and a 90-pip gold stop directly comparable — and removes sizing from the comparison entirely. |
| 5. Resolve honestly | Same-bar ambiguity resolves stop-first; unfilled limits expire; open trades are marked to market at a fixed horizon | Every tie-break is deliberately the pessimistic one. We'd rather understate a source than flatter it. |
| 6. Rank | Win rate, average R (expectancy) and profit factor over the last 50 resolved signals — minimum 10 before any rating shows | Expectancy, not win rate, decides the ranking. A 70%-win source that loses big on the 30% is not a good source. |
SOME RANK WELL.
MANY DON'T.
Here is the leaderboard as it actually appears — including the sources that score badly. We publish the negatives because a ranking that only ever flatters is an advertisement, not a measurement.
Each row carries an honest provenance line: "includes provider-published history · SentryQ-verified live since [date]". Backfilled history is scored the same way but we didn't witness it being posted, so we separate it from signals we recorded live. As live sample grows, the verified portion takes over. We'd rather show you a caveat than a cleaner number.
A SIGNAL DOESN'T
GET A FREE PASS.
This is where SentryQ diverges hardest from every copy bot on the market. A signal you approve is not executed blindly — it enters the same guard chain as an automated strategy trade.
If your net USD exposure is already at its cap, a USD signal is declined. If EURUSD already holds its maximum positions, another EURUSD signal is skipped. If the News Guard window is open, if the spread has blown out, if it's Friday evening, if margin is below your floor, if the daily loss limit has tripped — the signal is refused, and the reason is logged in your day summary alongside everything else.
Multi-target signals are handled properly: split into tranches that share one stop, counted once against your per-symbol cap so a legitimate three-TP signal isn't punished, and managed to breakeven once the group banks profit.
