Models defined per intent for C# and .NET, so one application can use several models easily, within one context or across several. For an IntentSharp catalog intent it sends the definition and the context you choose to a specialist service you wire in, waits until a specialist is provided, checks that the returned definition and specialist fit (a comparison in code by default; probe cases and an AI judge are opt-in), keeps the specialist's name and the returned intent for reuse, and runs the intent on that model through the normal IntentSharp pipeline, so every answer is still checked against your type and your rules. It never trains, tunes or changes a model and names no specialist provider: the service adapter, the store, the context sources, the fit checks and what happens after an offer does not fit are seams, and every setting is an option with a default. An offer names a provider configured on the client or brings its own, which the client reaches through a router by name only. A changed definition requests again, and keep-previous lets the specialist in use serve meanwhile. A document store keeps requests, tickets, fit reports and held specialists across restarts and processes; wait windows and an hourly request limit bound the exchange, and a guardrail checkpoint can refuse a request before anything is sent. A scripted specialist service and a contract kit test applications and adapters offline. Installs the matching IntentSharp and IntentSharp.Catalog packages. Commercial software, not open source: evaluate free for 30 days on development and test systems; after that, business use needs a paid license, and production use needs one from the first day (intentsharp.com/pricing).