Cases and evidence
Find real adoption records, field evidence, synthetic examples, and historical experiments without confusing one category for another.
AletheIA documents different kinds of proof. They should not be treated as interchangeable.
Field evidence is labeled Examples are not proof Guidance remains canonicalHow evidence is classified
| Type | What it is | What it can support | What it cannot prove |
|---|---|---|---|
| Canonical guidance | Contracts, ADRs, policies, and accepted definitions | Intended rules, boundaries, and vocabulary | That a rule worked in a real project |
| Field evidence | A source-backed real pilot, adoption record, or Work Slice closeout | What happened in that bounded context | Universal performance or a reusable control by itself |
| Synthetic example | An illustrative template, fixture, mock record, or walkthrough | How to use a structure or read an artifact | Real adoption, measured benefit, or enforcement |
| Historical experiment | A preserved trial or superseded path | Context for past decisions and learning | Current support or active direction |
Featured field evidence
Crisis Monitor pilot
The first real product adoption. It is a validation case, not a domain-specific requirement for every adopter.
AletheIA + Adaptive Skills dogfood
A real execution record projected as a read-only Resource Observatory signal without creating authority or success claims.
Resource-aware operations review
Consolidates what the bounded resource-aware work actually supports and where it stops.
Context graph decision
A decision record using controlled test data rather than assuming a tool should become general practice.
Closeouts and operation records
Closeouts capture what happened in a bounded operation: intent, scope, tools, evidence, decisions, remaining risk, and the next safe step. They are useful evidence when their source references and limits are clear.
Browse operation closeouts for preserved records. Treat them as contextual evidence, not an instruction manual or a generic performance benchmark.
How to use a case responsibly
Check the label
Confirm whether you are reading guidance, field evidence, a synthetic example, or historical material.
Read the boundary
Identify project context, scope, evidence source, missing data, and explicit stop conditions.
Transfer only what is justified
Reuse a pattern or question when it is portable. Do not carry over product-specific assumptions, synthetic numbers, or unstated permissions.
Create your own evidence
Apply guidance in a real bounded Work Slice and record the result before treating it as local proof.