A single inquiry is the basic unit of Shushu, and it can begin with either a question or a chart that has already been generated.

In Shushu, a chart is first calculated by the relevant shushu engine from valid inputs according to the rules of that system. Chat does not replace that calculation. Instead, it brings the chart, real-world facts, observations from other shushu systems, and the methods and texts that support interpretation back to the same question. Through follow-up, comparison, and calibration, these materials begin to illuminate one another, and a judgment gradually takes shape.

These elements play different roles: shushu provides observations, reality provides calibration, and methods and texts provide the basis for interpretation.

Different shushu systems select different structures of time, positional relations, symbols, and rules to build different models of a complex reality. They do not shrink the world until it fits completely inside a chart. Each selects certain relationships and makes them visible. They face the same reality, but approach it from different positions, scales, and relationships. What they reveal therefore overlaps in some places and diverges in others.

Shushu offers observation, not a substitute for reality. An observation still has to be interpreted and compared, then checked against real-world facts and the relevant methods and texts before it can enter a person’s understanding and contribute to a judgment.

1. Quick Start

1.1 Know the Entry Points

Shushu has five main entry points on the home page, each with a different role:

Fate book and Divination first establish the chart and provide the basic material for interpretation. For deeper interpretation, follow-up, calibration, or comparison across perspectives, move into Chat.

The interface currently lets you generate and view Bazi, Ziwei, Liuyao, and Meihua directly. Other shushu systems can be called selectively by a diviner according to the question at hand and the scale of observation required.

1.2 Choose the Right Starting Point

Where you begin depends on the question. If you want to understand your longer-term life structure, start with Fate book; if you are asking about one specific matter, start with Divination. If you already have a precise question and an existing profile, you can also go straight to Chat.

For a first-time user, the shortest complete path is:

Home → Generate a Fate book → Enter birth data → Generate profile and chart → Interpret the chart → Enter API Key → Open Chat

Selecting “Interpret this chart” from a Fate book or Divination result opens a new conversation with a fixed, uneditable context prefix that identifies the profile and shushu system used for that interpretation. You can send it as it is or add more information first; the system does not send anything automatically on your behalf. Once sent, the corresponding shushu engine is called again from the locked profile or case data, and its result is brought into the current conversation for further interpretation.

For divination, the original question, casting time, casting method, and original inputs together define the conditions under which that observation was formed. Divination follows the principle of one matter, one casting. The question arises and the casting takes place at a particular moment. What later happens in reality may change how we understand the original reading, but it cannot change the observation that was made at that moment.

For that reason, additional text in Chat may change the focus of interpretation, but it cannot rewrite the original divination case. Nor does the later time at which an interpretation takes place replace the original casting time.

You do not have to generate a chart first. You can also begin directly in Chat:

I’m thinking about changing jobs. Where should I start with this question?

The diviner will first understand the question and determine what information and forms of observation are needed. If a chart is required, it will ask for the inputs that the relevant shushu system actually needs before that chart can be established.

1.3 Understand the Chart

In traditional Chinese shushu, a chart is the calculated result produced when valid inputs enter a particular shushu model and are processed according to that model’s structure and rules. It is also the raw material for everything that follows.

It is first established according to the method itself. Only when the chart enters Chat is it brought back into a particular question, a real-world context, and the relevant theory.

It helps to distinguish four levels:

Take Bazi as an example. Bazi is a shushu model. A Bazi chart is the result produced by that model from a particular set of birth data. Interpretation begins only after the chart exists: its structures are read through theory, interpretive methods, and the real question being asked. A person forms a judgment by placing these observations back into reality and deciding how much confidence they deserve, which conditions still hold, and what must remain open.

A shushu model, a chart, an interpretation, and a judgment are not the same thing.

A chart does not become a different chart because someone later describes it differently. Our understanding of it—and the judgment we eventually form—can still be re-examined as facts, methods, sources, and questions change.

For a user, learning to calculate charts by hand is less important than gradually learning to distinguish what comes from the chart, what belongs to interpretation, what already has support in reality, and what remains only possible.

A chart must first be that particular chart. Only then can we ask how it should be interpreted.

1.4 Begin the Conversation

Once you have a question, or once a chart already exists, deeper interpretation and follow-up begin in Chat.

Before going further, distinguish three things: the diviner, the profile, and the question.

The diviner is responsible for understanding the question, asking follow-up questions, calling tools, organizing sources, and moving the interpretation forward. It does not alter a chart that has already been established.

Different diviners think through the same question in different ways: what they attend to, how they ask questions, which tools they call, how they organize evidence, and how they turn complex information into a current analysis.

Shushu currently has three diviners: Shushu, Ming, and Moon.

The profile determines the fixed subject of the inquiry when a conversation begins. The speaker is always the currently logged-in user; profile selection determines whether an existing profile is treated as that fixed subject and, if so, how that person relates to the speaker.

This choice remains fixed when the session begins. Later questions may involve other people, relationships, or events, and the diviner may consult another profile within the conversation, but the identity of the current speaker does not change.

The relationship among the three can be summarized simply:

The profile fixes the subject of the inquiry. The question determines what is being examined. The diviner understands the question, calls the necessary tools, and carries the interpretation and follow-up forward.

2. Calibrating Information

Shushu does not require elaborate professional prompts in order to go deep. You can begin with the simplest possible question and obtain a first useful observation.

But shushu calculates structures, while reality always contains information outside the chart. So the next task is not to collect more answers as quickly as possible, but to keep asking:

What is still missing from the current judgment? And where should that missing information come from?

2.1 Answer Follow-Up Questions

When the inquiry reaches a point where new information is needed, the diviner will ask for it.

There are roughly three kinds of cases:

1. The current interpretation lacks grounding in reality. More real-world information is needed because it could change the judgment.
2. The current chart does not cover the scale required by the question. We need to decide whether another shushu system is needed.
3. A new shushu system requires specific inputs before a valid chart can exist. Those inputs must first come from an identifiable source before the corresponding engine can be called.

The third case is especially important because, in some shushu systems, the inputs determine the chart directly. If an AI first develops an interpretive preference and then works backward from that preference to generate the necessary inputs, it may produce a chart that happens to support its original conclusion—and then use that chart to “verify” the interpretation.

The loop looks like this:

Interpretive tendency → interpretation influences the inputs → inputs produce a chart that supports the interpretation → the chart is treated as independent evidence that “verifies” the interpretation

On the surface, it may look as though “a second shushu system reached a similar conclusion.” In reality, the new observation has already been contaminated by the earlier interpretation and is no longer independent.

Shushu therefore places explicit gates around such critical inputs. A diviner can determine what is missing, ask the user for it, and call the relevant shushu engine. But the system does not allow AI to invent raw information that should come from the user, an existing profile, or another clearly identified source.

Another kind of follow-up occurs after a chart and an initial interpretation already exist. For example:

Have you two actually been in contact recently?

What is missing here is real-world information outside the chart, not a new input for generating one.

A shushu chart is a structured set of symbols—what traditional shushu calls xiang. Xiang can carry symbolic meaning, but symbolic meaning cannot automatically supply facts such as whether two people have actually met or whether a contract has already been signed.

The system can look for missing information, but it cannot speak on behalf of the source of that information.

The user can also decide in what order real-world facts enter the interpretation.

Sometimes you may want to ask: what can this chart show before I tell you what actually happened?

In that case, you can temporarily withhold certain facts, allow the diviner to complete a relatively independent first reading, and only then bring those facts back for calibration.

What matters is this: the system cannot supply information that was never provided and treat it as fact. And when new facts do enter, the interpretation must allow them to change the current judgment.

2.2 Add Facts

When you add information from reality, another distinction becomes important: what exactly counts as a fact?

“He hasn’t replied to me in two weeks.” — fact

“I think he’s avoiding me.” — judgment

“I hope this relationship can continue.” — expectation

All three are relevant to the question, but they do not carry the same weight once they enter the conversation.

Whether something enters the context as a fact, a judgment, or an expectation affects how the diviner interprets it in the next round and what role it plays in the current analysis.

A conversation can only work with the part of reality that has entered the inquiry so far. A user is free to select, omit, and organize their own account, but that account is not reality in its entirety. Facts that never enter the context—and cannot be obtained through an allowed source—remain unknown to the diviner.

The information already provided can help calibrate the present inquiry, but it does not become the whole truth simply by entering the conversation, nor does a more coherent—or more desirable—answer mean that reality has supplied more support.

2.3 Check for Bias

Interpretation always remains distinct from the chart itself.

The chart first provides structures and signals. An interpretation forms around them. As the conversation continues, later language begins to refer back to earlier interpretations. Even when no new chart, fact, or source enters the conversation, the account can still grow more complete.

But completeness is not new information. If new real-world facts appear later, they do not reach backward and change the old chart; they may, however, change our understanding of the existing chart and how much support remains for the judgment we originally formed.

In a long conversation, keep three things separate:

• Structures in the chart come from the shushu model and its rules of calculation.
• Interpretations of those structures come from the theory, method, and interpretive path currently being used.
• Judgments about reality are constrained by the chart, its interpretation, real-world facts, and other sources at the same time.

The xiang in a chart can support more than one real-world interpretation.

At any point, you can therefore ask the diviner to separate the current understanding again:

What comes directly from the chart?
What belongs to interpretation?
What already has support in reality?
What is still only possible?

The earliest interpretation should not be treated as permanently correct simply because it came first. What a long conversation needs to preserve is not the first story, but an interpretation that stays open to new facts.

Reality is always larger than what shushu can show.

2.4 Adjust the Path

Within the same conversation, a diviner can continue interpreting the current chart, ask for missing real-world information, verify methods and textual sources, or call another shushu system when doing so would genuinely add a new form of observation.

Choosing the next step does not need to begin by pursuing more—or more “accurate”—conclusions. First identify what is actually missing; once you know that, the next step becomes meaningful.

You can constrain the path directly:

Don’t add another shushu system yet. Read this chart in detail first.

Or:

If we add another shushu system, what information would it provide that we do not already have?

The next step is determined by what information is missing. If the same core question remains active and the materials already gathered still need to be compared, going deeper is often more valuable. A new method becomes useful only when it can provide information that the current material does not cover.

3. Cross-Checking

Reality is not made up of static elements. Its parts are constantly changing, gaining and losing strength, moving in different directions, and acting on one another. A person enters this movement at a particular time and place, with particular endowments, while living within a body, a family, a society, and a network of relationships.

This changing organization of interacting forces can be understood as an energy structure. When time, space, cycles, number, xiang, position, relationships, and change are organized together, a situation that has not yet become fully visible acquires a form that can be observed and reasoned through.

Different shushu models select different variables and scales from that dynamic structure, allowing a person to observe, compare, and form judgments from different angles.

3.1 Add Observations

The value of bringing multiple shushu systems together does not lie in having more answers. It lies in the fact that they organize relations in reality at different scales.

Changing systems should add more than another answer: it should introduce a new scale of observation or bring a different set of relationships into view.

These systems face the same reality, but they do not select the same relationships from it.

Some begin with the long-term structure of the individual. Others focus on an immediate event, a field of time and place, a complex unfolding situation, or the surrounding environment. Still others widen the scale toward longer historical cycles.

Different interpretive approaches within the same shushu system can be read together, but they still share the same underlying chart; what they add is another path of interpretation. Only when a new shushu system brings in genuinely new material does cross-checking begin to carry real weight. So rather than asking:

Which system should I add next?

ask:

What has the current system not observed? Does the new one genuinely change the scale of observation?

For example, suppose Bazi already reveals a long-term career structure. If the question then narrows to one specific opportunity in front of you now, an observation built around the particular event or its immediate conditions may add something genuinely new.

Once a second and third perspective enter, the question changes again: we are no longer asking only how many answers we have, but how observations from different shushu systems should be brought into the same judgment.

3.2 Check Independence

Once several shushu systems have produced results, first examine how independent they actually are. Only then ask how much support their agreement provides.

Suppose one inquiry produces:

Bazi: supports judgment A
Ziwei Doushu: supports judgment A
Liuyao: supports judgment A
Meihua Yishu: supports judgment B

That 3:1 cannot simply be converted into a conclusion.

Bazi and Ziwei Doushu may share the same birth data while organizing it through different structures. Liuyao and Meihua Yishu may share a set of reported numbers and a casting time while still placing different emphasis on how those inputs are interpreted.

Different shushu systems therefore have different degrees of independence. Su Shi wrote:

Seen from one side, it forms a range; from another, a peak. From far or near, from high or low, no view is quite the same.^[2]

Changing the observer’s position does not change Mount Lu itself. But approaching the same mountain from the side or the front, from near or far, from high or low, reveals different aspects of the same thing.

The Zhuangzi makes a related point:

At a finer scale, the larger whole cannot be fully grasped; at a broader scale, the finer details cannot be clearly seen.^[3]

The scale of observation determines what enters the field of view—and what remains temporarily invisible.

A change of scale is therefore not merely another way of saying the same thing; it may expose a different layer of relationships and information within reality. The same is true when several shushu systems are read together: faced with the same person or event, they may not occupy the same observational position or work at the same scale.

Only a genuinely different vantage point can bring information that the previous one did not provide.

Cross-checking therefore requires four questions:

If several judgments ultimately depend on the same critical premise, their shared support for judgment A remains limited. When observations with different sources and scales converge at a critical point, their agreement carries more weight—but even then, a majority result does not become fact.

Agreement is not the same as independence.

The same principle applies when comparing schools or interpretive approaches within a single shushu system. Different approaches based on the same underlying chart can check one another, but they should not be counted as separate pieces of independent support.

Likewise, if the inputs for a new shushu system have themselves been influenced by the conclusion of an earlier one, it becomes difficult to treat the new result as genuinely independent.

3.3 Handle Conflict

Disagreement between shushu systems also contains information.

A system focused on long-term structure may show that a particular relationship continues to hold an important place in a person’s life. A divination focused on the current event may show serious obstacles right now.

“Where this relationship sits in a longer life structure” and “whether it can move forward now” belong to different scales. Both can therefore be true at the same time.

When observations conflict, do not immediately let one override the other. First ask where the disagreement actually occurs.

If conflict remains after these conditions have been checked, it can remain part of the judgment.

Different signals from two valid shushu systems, each within its own limits, are themselves part of what we know. Some differences need explanation; others simply need to be preserved accurately because disagreement itself can expose differences in scale, input, assumptions, and interpretation that a smooth consensus might otherwise conceal.

The purpose of reading shushu systems together is not to make them all say the same thing in the end.

3.4 Verify the Grounds

Some disagreements are not between shushu systems at all; they concern the grounds on which an interpretation depends.

If the question is whether a chart was established under the right conditions, check the calculation conditions relevant to that method: time, calendar system, location, direction, and so on.

If the question is why the chart is being interpreted in a particular way, return to the relevant method, classical text, or commentary. Retrieve the relevant passage and, where necessary, read its original context.

When you want to examine a specific judgment, you can ask the diviner to separate the chart, the source, and the interpretation:

What specific structure in the chart supports this judgment?
If it depends on a method or classical source, separate the original source from the interpretation you are making now.

A retrieved text does not automatically become a conclusion. It becomes one piece of material—alongside the chart and real-world facts—that can inform a judgment.

Source and interpretation need to remain distinct because a statement in a classical text cannot simply be removed from its original context and applied directly to a contemporary situation.

4. Returning to Judgment

By this point, the conversation contains charts, real-world facts, observations from different shushu systems, and relevant methods and sources. Their agreements, repetitions, and conflicts have also been examined.

Once enough material has accumulated, collecting more answers is no longer the most important task. The real question becomes:

What judgment can these materials support now, and how should I decide what to do?

Charts, interpretations, real-world facts, and sources still retain their different roles.

Only by placing each one back within the conditions that make it valid can we see what now has support, what remains in motion, and what should stay open.

4.1 Form a Judgment

The first step in forming a judgment is to place every answer back within the conditions on which it depends.

You can ask directly:

Reorganize everything from this inquiry.

What is the best-supported judgment right now, and what supports it?
Which conditions are still changing?
What still cannot be determined?

No matter how certain a sentence sounds, certainty of expression does not create additional support.

Several results that share the same premise do not become several independent pieces of evidence simply because their wording agrees. An interpretation does not acquire new facts merely because it becomes more complete.

The task is therefore not to compress several shushu systems into one more confident sentence. It is to clarify how the existing materials relate to one another.

Only then can the information provided by shushu truly become part of a person’s understanding.

4.2 Understand Change

A judgment is the understanding a person forms from the observations, facts, sources, and circumstances currently available.

A judgment is not the future itself.

The Xici says:

The Yi enabled the sages to probe the deepest principles and discern the faintest signs of what was beginning to emerge.^[1]

The idea of ji points toward the first indications that appear before a matter has fully become manifest.

This gives us a practical boundary: relations that have already formed can be observed, and we can examine how changing conditions are acting now.

But what has not yet taken shape in reality cannot be written into certainty simply because a large language model has already produced an answer.

“I don’t know” can therefore mean at least three different things:

In the first two cases, more information can still be obtained. The last requires us to admit that reality is still unfolding.

Asking the same question again in different words, asking a different diviner, lengthening the same interpretation, or asking another large language model to answer again when no new material has entered—none of these creates new facts, conditions, observations, or sources.

Further inquiry becomes more meaningful only when genuinely new material appears.

New words are not new information. What has not yet taken shape in reality is not the same as missing information.

4.3 Know When to Act and When to Stop

Once a working judgment has formed, the question returns to reality:

Should you act now, wait, continue observing, or begin a new inquiry?

If new facts, observations, or sources still belong to the same core question, continuing the current conversation allows you to compare them directly with the existing chart and judgment. If the material already available is sufficient and reality has produced nothing new, you can stop and allow action—or waiting—to generate the next set of facts.

The future has not yet arrived. Knowing when to stop reasoning forward is itself part of judgment.

One hexagram in the Zhouyi is Gen, whose image is the mountain and whose meaning includes stopping and restraint. The Tuanzhuan says:

When the course of circumstances calls for restraint, hold steady; when it calls for action, move forward. Keep both action and stillness in accord with that course. The path then remains clear and sound. Gen, in its sense of “stopping,” means knowing when and where restraint is called for.^[4]

Stopping and acting are not simply opposite outcomes; they are two different rhythms of responding to circumstances. To stop at the right moment is itself a form of wisdom, and what matters is recognizing where present conditions ask us to hold and where they allow us to begin moving.

4.4 Make Room for the Unknown

Time continues, and reality continues to change. Yesterday’s judgment becomes part of the material available today, while what happened yesterday becomes part of the conditions we face today.

No matter how many rounds of reasoning Shushu carries out, it is still working with limited observations of structures, relationships, and change in reality. A human being still has to make sense of those observations.

Language cannot make reality happen, nor can reasoning by itself change the circumstances in which a person stands.

The choice, and the action, remain yours.

Time, environment, relationships, action, and new facts continue to alter the energy structure of reality. As those changes occur, they can be brought back into judgment, allowing the earlier understanding to be tested, revised, and carried forward.

The value of a Shushu conversation does not lie in eliminating the unknown, nor in predicting a “certain future.”

The inferences and predictions of every shushu model remain limited by its point of view. The analysis of trends and structures need only help a person understand more clearly where they stand, where the boundaries of their freedom lie, and what conditions shape the next step available to them.

That is enough. In the end, the question gradually shifts from:

“What is going to happen in my future?”

toward:

“How should I understand myself, and what should I choose now if I want to move toward the future I hope for?”

Know Tian, stand firmly within one’s limits and given nature, and the human comes to stand with Tian and Earth, each fulfilling its part.^[5]