Integrity Suite

The standards

The standards, all in one place.

Every standard, listed plainly so you can find the one you need. Each is independently adoptable and states its obligations, not an implementation, so conformance can be judged from the published document alone. They relate to each other as a set of active relationships, not a hierarchy, and each standard declares its own in its own text. Industry-specific standards, like the grants standards, are shown with their industry.

Precision-First Design Standard (PFDS)

The root the rest of the body grew from: precision and non-harming held as one move. Every other standard here is what that commitment requires, made precise for a particular kind of coordination.

Read more →

Dimensional Frame Language

Not a standard and not a domain instrument, but the shared vocabulary the standards are written in: the canonical terms a standard must use to carry its meaning intact, and the method for reaching them. It is what keeps a word from quietly meaning less than the check that rests on it. A plain-language walkthrough is on the Meaning page.

Read more →

Coordination Structural Integrity Suite (CSIS)

Ten standards specifying the structural conditions a coordination system must hold to. Seven are floors that must not be crossed, three are capacities that must be present. The suite ships with integration guides, patterns, paste-and-go AI prompts, and per-standard skills.

Read more →

CRAFT

The meta-standard for making a standard falsifiable and legible: six conditions a standard must satisfy to be checkable by an independent reader, plus the grammar for applying them to a domain. Its object is a standard as such, which is what makes it the one meta-standard rather than another domain standard.

Read more →

ORE

Origin, Reliability, Exposure. The ingestion boundary: what an evaluation chain is allowed to assume about the sources feeding it. Sources are graded as uncertainty rather than judged as quality, and every output exposes the grade profile of what it rests on, so nothing shaky is quietly laundered into a confident conclusion.

Read more →

STRUCK

The exit boundary, the mirror of ORE: what an output owes on the way out. Five obligations any evidence-grade result must carry on its face, so the party receiving it can assess it from the output alone. Support graded, Traced to origin, Refutation stated, Unjudged worth, Contest kept, Known on its face.

Read more →

WALKRI

Working Architecture for Legible, Knowable, Reliable Instrumentation. Epistemic quality at the point of data capture. Its unit is the instrument: any single place structured data is captured, defined so the data it produces is valid, consistent, provenance-aware, and reusable. Output aligns with Croissant, FAIR, and W3C PROV.

Read more →

TRAINS

Candidate

Temporal Representation And Interchange Notation Standard. The notation layer: a domain-neutral way to write down a chain's temporal profile, how fast it acts, how quickly its readings go stale, whether and for how long its actions can be undone, its cadence, and what its claims say about time. It represents and renders no verdict.

TIMES

Candidate

Temporal Integrity Measurement and Evaluation Standard. The judging layer: it reads a profile written in TRAINS and evaluates it against the corners where a reading correct at its source is consumed after it has gone stale, the failure a point-of-entry check never reaches. TIMES is to TRAINS as a judge is to a representation.

Free to adopt, with no string back to the author.

CSIS is published under Apache 2.0; the rest are CC0. The specification, the conformance tools, and the schemas are the whole of what you need to implement any of them.