# Ordo

> Ordo is Echelon Foundry's State-Directed Engineering methodology. It makes state, evidence, legal actions, obligations, uncertainty, and external-effect outcomes explicit so humans and AI agents can make bounded changes without reconstructing the system's meaning from incidental code.

Every quantitative claim on this site is backed by a record in an evidence manifest that names the artifact, the repository and the commit it came from, together with a statement of what the number does not show. Read the manifest first if you want the numbers without the prose.

## Machine-readable

- [Evidence manifest](https://ordo.echelonfoundry.com/data/evidence.json): every experiment, metric, claim and source, as JSON.
- [Glossary](https://ordo.echelonfoundry.com/data/glossary.json): defined terms, as JSON.
- [Page index](https://ordo.echelonfoundry.com/data/index.json): every page with its section and summary, as JSON.

## Concepts

- [Core concepts](https://ordo.echelonfoundry.com/concepts/): The nine ideas the methodology is assembled from, each with its evidence and its common failure mode.
- [Explicit domain state](https://ordo.echelonfoundry.com/concepts/state/): A named state instead of a combination of flags that a reader has to decode.
- [Legal transitions](https://ordo.echelonfoundry.com/concepts/transitions/): Which changes the model permits, declared once, enforced everywhere.
- [Capabilities](https://ordo.echelonfoundry.com/concepts/capabilities/): What an actor may do next, derived from the rules rather than maintained beside them.
- [Obligations, effects and uncertainty](https://ordo.echelonfoundry.com/concepts/obligations/): What is still owed, what leaves the system, and what is honestly not known yet.
- [Boundaries and authority](https://ordo.echelonfoundry.com/concepts/boundaries/): Where a constrained value becomes an open one, and what three of the experiments measured there.
- [Structural locality](https://ordo.echelonfoundry.com/concepts/locality/): A feature's parts stay together; the authority that decides its rules stays singular.
- [Navigation and context](https://ordo.echelonfoundry.com/concepts/navigation/): Deterministic routing to the right semantic area, and honest diagnostics when that fails.
- [Heterogeneous verification](https://ordo.echelonfoundry.com/concepts/verification/): Several mechanisms that fail differently, because no single one covers the failure set.
- [Evidence and traceability](https://ordo.echelonfoundry.com/concepts/evidence/): What was measured, how, and what must never be concluded from it.

## Evidence

- [Controlled Experiment 1 — boundary amplification on an unhardened architecture](https://ordo.echelonfoundry.com/evidence/helixnote-1/): How much of a new domain case's required propagation work a compiler can find on its own, and what happens to the part it cannot.
- [Controlled Experiment 2 — boundary authority hardening](https://ordo.echelonfoundry.com/evidence/helixnote-2/): Whether the blind spots Experiment 1 found can be closed mechanically, and what closing them does and does not change.
- [Controlled Experiment 3 — agent cost under hardening (corrected)](https://ordo.echelonfoundry.com/evidence/helixnote-3/): A paired A/B trial in which two isolated agents implemented the same frozen mission against the baseline and hardened architectures. The result that most complicates the programme's own thesis.
- [Greenfield Trial 1 — instrumented construction from frozen requirements](https://ordo.echelonfoundry.com/evidence/greenfield-1/): A time-entry system built from a frozen requirements package, with development-metrics instrumentation established before any implementation code was written.
- [Greenfield Experiment 2 — construction with no method authority in the repository](https://ordo.echelonfoundry.com/evidence/greenfield-2a/): The same frozen requirements, built in an empty repository that contained no architecture rules, no method documents and no conventions of any kind.
- [Greenfield Experiment 2B — independent replication and convergence comparison](https://ordo.echelonfoundry.com/evidence/greenfield-2b/): A second agent built the same requirements without seeing the first build, added an independent verification phase, and only then compared the two. The strongest single result on this site comes from that phase.
- [C# state-system trial — the same domain, structured explicitly](https://ordo.echelonfoundry.com/evidence/te-csharp-state/): The full time-entry domain built in C# against an explicit state model, reaching a first clean pass with no domain defects and no rework.
- [F# state-system trial — same requirements, same model, different language](https://ordo.echelonfoundry.com/evidence/te-fsharp-state/): The C# trial's requirements and state model reused verbatim and implemented in F#, holding the specification fixed so the comparison isolates language.
- [Conventional C# rebuild — ordinary practice, requirement by requirement](https://ordo.echelonfoundry.com/evidence/te-conventional-csharp/): The same domain rebuilt the way a great deal of production C# is actually written, with every mistake recorded live and classified by what caught it.
- [Conventional JavaScript rebuild — the same domain with no type checker at all](https://ordo.echelonfoundry.com/evidence/te-conventional-js/): The identical requirement set built a third time in JavaScript, to see what changes when the language offers no static checking whatsoever.
- [Controlled baseline — everything held constant except explicit domain types](https://ordo.echelonfoundry.com/evidence/te-controlled-baseline/): The most tightly controlled comparison in this evidence base, and the one that contradicts the easy conclusion. It is published because it did.

## Home

- [Ordo | State-Directed Engineering](https://ordo.echelonfoundry.com/): What Ordo is, how it works, and what it changes for human and AI-assisted engineering.

## Reference

- [Ordo, SDE, ROS and what is not here](https://ordo.echelonfoundry.com/about/): Naming, scope, and an explicit account of what was looked for and not found.
- [Glossary](https://ordo.echelonfoundry.com/glossary/): Every defined term, with its origin and its source document.

## Research

- [Research status](https://ordo.echelonfoundry.com/research/): The full map of propositions — supported, provisional, contradicted and open.
- [Evidence](https://ordo.echelonfoundry.com/evidence/): Every experiment behind Ordo, with provenance for every number.

## Start here

- [Why Ordo exists](https://ordo.echelonfoundry.com/why/): The measured problem in AI-assisted engineering, and what Ordo does about it.
- [How Ordo works](https://ordo.echelonfoundry.com/how-it-works/): The Ordo decision loop, four-tier responsibility model, and engineering method.
- [Results](https://ordo.echelonfoundry.com/results/): Every recorded trial, organised by what it concluded.
- [Ordo and conventional development](https://ordo.echelonfoundry.com/comparison/): Where the two approaches actually differ, and where they measurably do not.
- [Examples](https://ordo.echelonfoundry.com/examples/): Worked examples, each drawn from a real experiment rather than invented to illustrate.
- [Ordo in C#](https://ordo.echelonfoundry.com/csharp/): A complete C# walkthrough showing how Ordo's current semantics map into ordinary .NET code.
- [Adoption](https://ordo.echelonfoundry.com/adoption/): Four starting points, with what each one actually costs and what the evidence does not cover.

## Experiments

- [Controlled Experiment 1 — boundary amplification on an unhardened architecture](https://ordo.echelonfoundry.com/evidence/helixnote-1/): How much of a new domain case's required propagation work a compiler can find on its own, and what happens to the part it cannot.
- [Controlled Experiment 2 — boundary authority hardening](https://ordo.echelonfoundry.com/evidence/helixnote-2/): Whether the blind spots Experiment 1 found can be closed mechanically, and what closing them does and does not change.
- [Controlled Experiment 3 — agent cost under hardening (corrected)](https://ordo.echelonfoundry.com/evidence/helixnote-3/): A paired A/B trial in which two isolated agents implemented the same frozen mission against the baseline and hardened architectures. The result that most complicates the programme's own thesis.
- [Greenfield Trial 1 — instrumented construction from frozen requirements](https://ordo.echelonfoundry.com/evidence/greenfield-1/): A time-entry system built from a frozen requirements package, with development-metrics instrumentation established before any implementation code was written.
- [Greenfield Experiment 2 — construction with no method authority in the repository](https://ordo.echelonfoundry.com/evidence/greenfield-2a/): The same frozen requirements, built in an empty repository that contained no architecture rules, no method documents and no conventions of any kind.
- [Greenfield Experiment 2B — independent replication and convergence comparison](https://ordo.echelonfoundry.com/evidence/greenfield-2b/): A second agent built the same requirements without seeing the first build, added an independent verification phase, and only then compared the two. The strongest single result on this site comes from that phase.
- [C# state-system trial — the same domain, structured explicitly](https://ordo.echelonfoundry.com/evidence/te-csharp-state/): The full time-entry domain built in C# against an explicit state model, reaching a first clean pass with no domain defects and no rework.
- [F# state-system trial — same requirements, same model, different language](https://ordo.echelonfoundry.com/evidence/te-fsharp-state/): The C# trial's requirements and state model reused verbatim and implemented in F#, holding the specification fixed so the comparison isolates language.
- [Conventional C# rebuild — ordinary practice, requirement by requirement](https://ordo.echelonfoundry.com/evidence/te-conventional-csharp/): The same domain rebuilt the way a great deal of production C# is actually written, with every mistake recorded live and classified by what caught it.
- [Conventional JavaScript rebuild — the same domain with no type checker at all](https://ordo.echelonfoundry.com/evidence/te-conventional-js/): The identical requirement set built a third time in JavaScript, to see what changes when the language offers no static checking whatsoever.
- [Controlled baseline — everything held constant except explicit domain types](https://ordo.echelonfoundry.com/evidence/te-controlled-baseline/): The most tightly controlled comparison in this evidence base, and the one that contradicts the easy conclusion. It is published because it did.

## What this site will not tell you

No claim here is presented as generally established unless the evidence manifest records it as supported. Propositions this research has refuted are published alongside the ones it supports, with the same prominence. Where telemetry was never captured, the site says so rather than estimating.