Technology
How it is built.
An operome, your rulebook compiled so that a machine executes it exactly, is built in two moves. Compilation turns your governing documents into typed facts, computed values and rules, each traced to its clause. The evaluator then runs those rules over the facts of a case and returns the verdict with its trace, the same way every time. The APIs, the screens and the deployment are arranged around those two moves.
Compilation
The source text is marked up clause by clause. Each clause yields facts (typed), computed values (derived from facts) and rules (permitted, required, blocked, routed, escalated). Each element keeps the address of the clause it came from, so the trace is built into the file rather than written up afterwards.
The result is a typed, validated document that both the evaluator and the screens read. Before it goes live, a reviewer checks each rule against its clause, and a test run proves that the operome gives the expected result for every delivered test case.
Evaluator
The evaluator gives the same verdict for the same facts, every time. Facts are true, false or unknown. An unknown stays unknown rather than being guessed, so a missing fact shows up as a question and never as a permission. For machine actions the check fails closed: positive authorisation is required before anything is allowed.
Every verdict carries its trace, the rules that fired and the clauses behind them, and a hash over the run. An auditor recomputes the decision from the inputs instead of trusting a log. No model sits in the decision path.
Access by MCP, REST and CLI
- MCP. The operome appears to an AI assistant as tools: list sections, get a section, get the schema, evaluate, check a machine action, get the master rule, search. The StGB Operome answers this way now at https://operome.dev/mcp. Add it to your assistant.
- REST. The same check for your systems: send the facts, receive the verdict and the trace.
- CLI. The same check from a terminal or a CI job, for engineers and for pipelines that must refuse to deploy what the rules refuse.
Security and deployment
The Operome Runtime runs on your servers or EU-hosted and is licensed per operome and per environment. The operome is a file that travels with your deployment. The runtime needs nothing from us at decision time and sends nothing out; the facts stay where they are evaluated.
Who sees your data. The runtime needs nothing from us at decision time and sends nothing out. For client support we need to be able to access your data, so that possibility exists; our standard setup does not use it, and only a very limited number of our team members have access. If you use our deterministic rule sets with AI, the AI model of your choice reads your data too.
The public rule sets are open. The StGB rule set, its schemas and the method are at https://github.com/SynapseLayer-ai/stgb-operome under Apache-2.0, so you can read the method before you buy it.