BTC Orderflow Signals on Telegram: How the Buildix Signals Channel Works

On September 18, the day the channel went live, the Buildix entry engine had been checking BTC on Hyperliquid every minute since midnight UTC. By launch time it had flagged exactly one grade A setup, at 07:05 UTC. One. That is the idea behind the new public Telegram channel: BTC orderflow signals that show up rarely, in real time, with the reason attached.

The channel is called Buildix Signals and it lives at t.me/buildixsignals. It is free to join, it is fully automated, and nobody edits or curates what goes out. Here is exactly what it posts, what it leaves out, and how the engine behind it decides.

What the Channel Posts, and What It Leaves Out

Every post is a single BTC entry signal, grade A only, on the 5 minute or 15 minute chart. The message carries the side, long or short, the timeframe, the bar close time in UTC, the BTC price at the moment of the signal, and a plain sentence listing which orderflow conditions fired.

Posts go out the moment the engine flags the bar. No delay, no batching, no end of day digest. The same engine draws the entry markers on the BTC deep view chart, so the channel and the chart tell the same story.

What you will not find: stop loss, take profit, targets, position size, or leverage. The channel is a data alert, not a trade plan. It tells you that something structurally interesting just happened in BTC orderflow and why. What you do with that is your call.

Lower grades and other pairs are also out of scope. Grade B and C setups, and the rest of the top Hyperliquid pairs the engine covers, stay inside the product.

How the Engine Builds BTC Orderflow Signals: Six Checks

The engine is a confluence model. It runs six independent checks on each closed bar and only emits a signal when enough of them agree.

The anchor is a liquidity grab: a wick through a qualified level, such as a swing high or low, a cluster of equal highs or lows, or a high volume node from the volume profile, followed by a close back inside. The size of the wick is normalized by ATR with both a floor and a ceiling, so tiny pokes and genuine breakouts are both rejected. If the anchor does not clear its floor, no signal exists.

Five more checks then vote. Flow rejection asks whether the signed delta of the bar opposes the grab, meaning aggressive sellers pushed through the low and got nowhere. Absorption looks for high volume with almost no price progress at the level, the classic effort versus result mismatch. CVD divergence checks whether price made a new extreme while cumulative volume delta refused to confirm it.

Structure shift asks whether the next one or two bars hold the reclaim and push past the midpoint. Level confluence checks whether the grabbed level overlaps with other reference levels such as VWAP or the point of control.

At least three of those five have to agree, and the composite score has to reach 0.6, before a marker exists at all. The score then maps to a grade: A, B, or C. The channel only carries A.

Why Grade A Is Rare: The Regime Veto

On top of the six checks sits a regime filter with a hard veto. The rule is simple: never fade a strong trend unless absorption is present. A clean looking sweep of the highs in the middle of a one way move gets blocked, no matter how many other boxes it ticks. Vetoed signals never reach the channel.

The result is a low posting frequency by design. Some days produce one grade A signal on BTC. Some produce none. A feed that posts forty times a day turns the signal into the noise. A feed that stays silent until liquidity grab, flow, and structure line up is closer to how orderflow trading in crypto actually works: long stretches of nothing, then a short window where the tape says something clear.

All of it is computed on Hyperliquid data: 5 minute bars and the trade feed for signed delta, with the 15 minute view aggregated from the same bars. No sentiment feeds, no discretionary override, no human in the loop.

Every Signal Gets a Public Follow Up

Real time posting has a cost. Some signals will fail, and a public channel cannot quietly delete them. So it does the opposite.

Four hours after each signal, the bot replies to its own post with the current BTC price and the percentage move relative to the signal direction, whether that move is in favor or against. Winners and losers get the same treatment. Every Sunday at 18:00 UTC a weekly recap follows: how many signals fired, how many were in favor after four hours, the average move, the best and the worst.

Telegram timestamps every message, and the follow ups are replies to the originals, so anyone can scroll back and audit the history. That is the difference between a signal feed and a track record.

Where the Full Trade Plan Lives

The channel is the public slice. Inside Buildix the same engine covers the top 12 Hyperliquid pairs by 24 hour volume, on both timeframes, across all three grades. Each marker on the chart carries the complete plan: entry, stop beyond the wick with an ATR buffer, two targets, and R multiples, plus a checklist showing which of the six checks fired. Private alerts by Telegram or email, filtered by pair, grade, and timeframe, are part of the paid plans at buildix.trade/pricing.

To see the context around a channel post, open buildix.trade/pair/BTC when it lands. CVD, order book imbalance, the volume profile, and the signal marker sit on the same screen, so you can judge the setup yourself instead of taking a bot's word for it.

Join at t.me/buildixsignals and expect silence most of the day. When the channel speaks, the message tells you what BTC orderflow just did and why the engine cared. None of it is financial advice, and none of it asks for blind trust: the follow ups are there so the record speaks for itself.