The most basic boundary
you can put around an AI surface.
Three edges. One repository. Downloadable as one Markdown file. Drop it into a new GitHub repo and your AI surface has guardrails it did not have twenty minutes earlier.
Download SOVEREIGN-STARTER.md
Read the landing
The operator on one side of the repository, the objective on the other. The triangle in between binds every token of the model's processing to the axis between them — EVE Enterprise Harness at entry, canon at exit, observation registry logging every attempt to leave that axis.
The three edges
1.EVE Enterprise Harness
How the AI surface behaves before it does anything. The read-first rule. The one edge every session starts on.
2.Canon
How the artifact leaves the repository. The output shape. The edge the work travels along on the way out.
3.Observation registry
How the truth of what happened is preserved. The record of what was done wrong so it can be done right. The edge that closes the triangle.
Take the EVE Enterprise Harness away and the surface drifts. Take the canon away and the artifact leaks. Take the observation registry away and the other two become preferences instead of rules.
The thread the build inherits
The three edges govern the machine. The work it helps you produce needs a shape too. Pick any line of a build — a screen, a configuration, an interface, a test — and you should be able to walk it back to the decision and the client source that justified it. Part VII of the starter gives that record three layers.
1.Evidence
What the client gave us, kept with its source. Sources, needs, requirements.
2.Decision
What was chosen from it, by whom, and why. Decisions, assumptions, open questions.
3.Handoff
What the build receives and how done is proven. Blueprint, build items, acceptance.
The AI extracts and proposes. It never approves. The functional consultant owns the meaning, the client approves the decisions, and the sovereign build starts from a record it can trace instead of documents it has to reinterpret.
Why bother
The AI surface is a web browser. Nothing you say to it is permanent. Nothing you reason inside it comes back to you once the session closes.
Facebook is beginning to give people their history back. AI surfaces are not promising that yet. Until they do, the only way to keep what you think is to keep it somewhere you hold. You hold the repository. That is the whole idea. Everything else is mechanics.
This scaffold is not the sophisticated version. It is not the enterprise version. It is not the controlled EVEglyphDesign canon. It is the minimum shape that still works. More guardrails can be added on top later. The three edges are the floor.
How this connects to the doctrine
SAPfans.io asks for the right to add: to mirror the ledger read-only into a store the customer owns, and to build on top of it. The pattern is provider-neutral — fork the repo, point your AI at the record, build one twin. That step assumes you already have a repository your AI reads and writes against, and an operating shape that keeps its output honest. If you do not, the pattern does not land.
Sovereign Start is the floor beneath the pattern. It gives a first-time practitioner a repository, an EVE Enterprise Harness, an output canon, and an observation registry in one file — enough shape that an AI surface can be pointed at their record without producing toilet paper. Once the floor is in, the pattern becomes usable.
Download the file
One Markdown file: SOVEREIGN-STARTER.md. Save it as README.md in the root of a new GitHub repository.
Point your AI at it
Perplexity, Claude, ChatGPT, Gemini — whichever surface reads your repository. Ask it to read the file and follow the setup in §0.
Take the pattern
Now go to the Pattern. Fork the repo, mirror the record you care about read-only, build one twin. The floor is in.
Provenance
- Landing: eveglyphdesign.github.io/sovereign-starter
- Repository: EVEglyphDesign/sovereign-starter
- Source canon: EVEglyphDesign/eve-glyph-boot-contract
- License: practitioner-use notice