Skip to main content

Inner/Outer Knowledge Extension Model

What should everyone be free to learn, and what must remain specific to the system that makes the lesson repeatable?

The inner/outer knowledge extension model pairs two spaces with different jobs:

  • the outer space teaches WHAT and WHY in public;
  • the inner space controls HOW and WHEN in a particular operating system.

The boundary is not secrecy for its own sake. It is a way to give ideas away without confusing an explanation with an implementation, and to keep operational evidence useful by leaving it with the system that can interpret it.

Outer Space​

Public knowledge should help an unfamiliar reader understand a concept, test a claim, make a better decision, or try a method. It earns trust through usefulness before asking for attention, adoption, or payment.

The outer space contains:

  • definitions and distinctions;
  • first-principles arguments;
  • reusable methods and instruments;
  • examples, counterexamples, and failure modes;
  • public evidence and honest uncertainty;
  • links that lead to the next useful idea.
  • portable educational skills under .agents/skills/** that let a reader apply published doctrine without importing private operating state.

Education is free for the willing to learn. A reader can realize the value in their own time, adapt the reasoning to their own situation, and decide whether the method deserves further commitment.

Inner Space​

An idea becomes dependable only when a real system can execute, measure, and correct it. The inner space contains that contextual machinery: current state, responsibilities, controls, tests, evidence, exceptions, recovery, and the next decision.

Those details are not universal teaching. They may depend on private data, current architecture, security boundaries, customer commitments, or changing operating conditions. Publishing them can expose risk, but even harmless details can mislead: a command or threshold copied without its system may look like doctrine when it is only one implementation.

The inner counterpart to this page is therefore an operational boundary contract. Its job is to prove HOW and WHEN, enforce safe publication, and cite this public page for WHY. This page does not reproduce that machinery.

Compounding Loop​

Either space alone reaches a limit.

Public ideas without operating proof can become attractive but fragile. Private controls without public reasoning can become efficient but isolated: outsiders cannot learn from them, insiders forget why they exist, and trust depends on claims nobody can inspect.

Together they form a learning loop:

  1. Public reasoning gives a system a clear purpose and shared language.
  2. Private practice tests whether the reasoning survives contact with reality.
  3. Evidence changes the method, its bounds, or confidence.
  4. Only the reusable lesson returns to public education.
  5. More readers test, critique, teach, and extend the idea.

The public commons grows. The operating system improves. Neither needs to pretend it owns the other’s job.

Enforcement Advantage​

A durable advantage need not come from hiding the worldview. It can come from the accumulated ability to practice it consistently: instrumentation, feedback, coordination, judgment, recovery, and standards that improve after every run.

Giving away the WHAT and WHY can strengthen that advantage:

  • useful education attracts better questions and collaborators;
  • clear doctrine reduces explanation cost;
  • public criticism reveals weak claims;
  • shared language improves coordination;
  • readers arrive with more informed expectations;
  • demonstrated generosity earns trust before a transaction.

The lesson is not “publish everything.” It is “do not mistake obscurity for a moat.” Keep sensitive and context-bound operations private. Teach the reasoning that helps others create value.

Promotion Test​

Promote a lesson only when all five tests pass.

1. Reusable​

Would the lesson remain useful outside the incident, customer, project, or tool that produced it?

2. Evidenced​

Can you distinguish what was demonstrated, what is inferred, and what remains an open question?

3. Safe​

Can the lesson stand alone without private identities, data, credentials, commitments, vulnerabilities, or operational state?

4. Teachable​

Can an unfamiliar reader understand why it matters, apply a decision test, and recognize where it fails?

5. Placed​

Does an existing public page own the job, or is a new page necessary and linked from a clear parent route?

If any test fails, keep learning privately. “Not public yet” can mean immature, context-bound, unsafe, duplicated, or simply not useful enough.

Worked contrast​

Suppose a team repeatedly loses time because names drift across documents, software, and handoffs.

The public lesson is the Name Integrity Stack: establish what exists, how it is known, where it belongs, what it is called, and why the name improves action.

The private implementation may include repository scanners, migration receipts, ownership rules, and baseline data. Those controls prove that one system follows the lesson. They are not required for a reader to understand or apply the model, and they should not become universal instructions merely because they work in one environment.

Dream-local educational skills are legitimate outer content when they remain portable and cite published doctrine. Secrets, private state, hooks, generators, customer evidence, and operational control-plane scripts remain in Stackmates.

The two sides cite one another by role: the public page explains the principle; the private contract enforces the local application.

Failure modes​

  • Public manual: a teaching page becomes a list of private commands, paths, states, and exceptions.
  • Private philosophy: an operating contract repeats the whole public argument instead of citing it and proving execution.
  • Proof stretching: a local success is presented as a universal law.
  • Premature promotion: raw notes or aspiration are published before a reusable lesson exists.
  • Secrecy theatre: useful reasoning is withheld even though the real advantage comes from practice and feedback.
  • Open-washing: public language promises transparency while material risks or uncertainty are hidden.

Decision rule​

Ask two questions:

  1. What understanding would help another person create value safely?
  2. What context-bound machinery is required to make that understanding repeatable here?

Teach the first. Govern the second. Connect them with evidence, not duplicated authority.

Context​

Questions​

  • Which part of this lesson is universal, and which part works only because of its current operating context?
  • What evidence would let another reader trust or falsify the public claim?
  • Can you teach the model back without naming a private command, path, customer, or project state?
  • What must remain governed for the lesson to work repeatedly?

Teach-back: explain the boundary as “share the reasoning; govern the context-bound execution,” then name one example of each side.

Changes my mind: evidence that publishing context-bound operating detail improves safe reuse without confusion, leakage, or authority drift would narrow this boundary.

Next question: which proven private lesson is now mature enough to teach in public without carrying its implementation state with it?