Inside the structure pattern Keystone is built to find

What it is, and what standalone means

Keystone is a standalone scanner for structure on XAUUSD. Standalone is the operative word. It is not part of a suite of confirmations, it does not combine readings from other engines, and it is not trying to produce a complete trading method. It looks for one thing: a 1-2-3 pattern that resolves into an A-B-C leg with a break of structure.

That is the entire specification, and the usefulness comes from the restriction. A scanner that looks for one well defined sequence can be evaluated honestly, because you can see whether the sequence is there or not. A scanner that blends many conditions produces output nobody can audit, and when it is wrong there is no way to tell which part was wrong. The trade off is coverage. Keystone is silent through everything that is not this pattern, which is most of the time, and silence from a narrow tool is not a malfunction.

The 1-2-3 sequence

A 1-2-3 is the simplest description of a trend that has stopped working. Taking it in a declining market, the three points are:

• Point 1: the extreme of the existing move, the lowest point reached before anything changes.
• Point 2: the high of the reaction away from that extreme. Price bounces and then stalls, and this is where it stalls.
• Point 3: the low of the pullback after that reaction, which fails to make a new extreme beneath point 1.

The information is in point 3. The market tried to continue and did not get there. Nothing has been proven at that moment, since markets fail to make new extremes constantly and then go on to make them later, but a condition now exists that can be tested: if price takes out point 2, the sequence of lower highs has been broken. In an advancing market everything inverts, with point 1 as the high and point 3 as the lower high that fails. The underlying grammar of swing points is covered in how structure is read on gold.

How it becomes an A-B-C leg

Once the 1-2-3 is in place, the next phase is labelled A-B-C, and it describes how the new direction actually unfolds. A is the first impulsive leg away from point 3, the move that does the work of breaking the previous pattern. B is the retracement against it, where the market pulls back and tests whether the new direction has genuine participation. C is the continuation leg that follows.

The reason this matters is that it separates two different things a trader can wait for. The 1-2-3 identifies a potential turn. The A-B-C describes the behaviour that would be consistent with that turn being real. A move that produces A and then never manages C has told you something, because the pullback at B absorbed everything the new direction had. The sequence is also where the concept of a protected level becomes concrete, since point 3 is the low that the new direction is implicitly defending, a notion developed in the protected swing idea.

The break of structure, and its ambiguity

The break of structure is the event that ties the two halves together. It is the moment price takes out the swing point that had been holding the previous pattern in place, which in the declining example is point 2. Before the break you have a hypothesis. After it you have a changed structure.

Here is where honesty is required, because the definition contains a real choice. Does a wick through the level count, or do you require a candle body to close beyond it. Both conventions are used, neither is correct in any absolute sense, and the two will disagree several times a week on gold. A wick based definition registers the break earlier and catches more that fails. A close based definition is slower and misses some genuine turns entirely. Choosing one and applying it consistently matters far more than which one you choose, and the trade offs are set out in how breaks of structure are distinguished from sweeps.

Reading it on the chart

A practical order of operations. First find the extreme the move came from, which is point 1, and be strict about it rather than choosing a convenient candle. Then identify the reaction high at point 2, because that level is the one that will decide whether anything happens. Then check whether the pullback low at point 3 genuinely failed to exceed point 1. If it did not fail, there is no pattern and you are looking at a continuation.

From there the work is waiting. The level at point 2 is fixed, your definition of a break is fixed, and nothing requires a decision until price reaches it. This is the part most people skip, because sitting with a marked level and doing nothing feels like a wasted session. On the live gold chart you can watch price approach that level in real time, which is the one part of this exercise that requires no interpretation at all. The scanner identifies the sequence. It does not decide whether the context around it deserves a position.

Where structural reading goes wrong

Three failure modes, and the first is the one nobody mentions. Swing points are not fixed while a chart is still forming. What looks like point 3 can be redefined as new candles print, and a pattern that appears clean in hindsight was ambiguous for most of the time it was developing. Any structural tool inherits this, and it is the reason backtested structure almost always looks tidier than live structure.

Second, a sweep looks exactly like a break until it is over. Price runs through point 2, collects resting orders above it, and returns below within a candle or two. The structure was never actually broken, but a wick based definition already recorded it. Third, the pattern is scale dependent. The same price action contains 1-2-3 sequences on several timeframes at once, pointing in different directions, and nothing about the pattern tells you which one is the relevant one. That choice is yours and it is where most of the real disagreement lives.

What it does not provide

Worth stating directly. Keystone identifies a structural sequence. It does not tell you to trade, it says nothing about how far a move might travel, it does not size a position, and it makes no claim about how often the pattern resolves one way or the other. It has no view on whether conditions are suitable, so a textbook sequence can appear in the middle of a session you should not be trading and the pattern will look exactly as good as one that appears in a clean trend.

It also carries no information about why the structure changed. A break driven by a scheduled release and a break that develops slowly through a quiet session produce the same labels and behave differently afterwards. The scanner cannot distinguish them, and expecting it to is expecting a structural tool to be a macro tool. Related reading on how a shift in structure is interpreted sits in the market structure shift concept.

Using a narrow tool well

The discipline that suits Keystone is simple and unexciting. Decide in advance which timeframe you care about, so that competing sequences on other frames do not become an argument you have with yourself mid session. Fix your definition of a break and write it down, because the version you prefer tends to drift towards whichever one would have been right last time. Then treat an identified sequence as a reason to pay attention rather than a reason to act.

The honest summary is that this tool removes one kind of work, which is spotting the sequence, and leaves all the harder work untouched. Judging relevance, judging conditions, deciding size and accepting that a clean pattern can fail completely are still yours. A narrow scanner used with that understanding is genuinely useful. The same scanner treated as a signal generator will produce exactly the experience you would expect from trading one pattern without regard for anything else.