Architecture
Overview
AI Foundation is a shared foundation layer:
Consuming Extension Code
|
v
AiServiceInterface
|
v
AdapterRegistry -> Provider adapters
|
v
Provider APIs
Parallel support:
AiStatisticsService -> OpenAiOrganizationUsageService -> OpenAI Usage API
HttpAuthUtility -> Protected URL fetching with optional Basic Auth
AI Prompts (T3AI)
Management UI for T3AI global and sidebar prompts lives in AI Foundation → AI Prompts.
Built-in defaults are defined in ns_t3ai (PromptContractRegistry); the database stores custom overrides only.
See PromptManagementProvidersAndFeatures.md and PromptsSystemReference.md in this directory for routes, UI behaviour, and class reference.
Main components
Request orchestration:
AiServiceInterfaceandAiServiceProvider adapters:
AdapterRegistryandAdapterInterfaceimplementationsStatistics processing:
AiStatisticsServiceandOpenAiOrganizationUsageServiceEngine configuration filtering:
AiEngineConfigurationUtility and environment helpers:
AiUniverseUtilityHelperHTTP auth helper:
HttpAuthUtility
Configuration model
Runtime behavior is mostly driven by extension configuration keys from
ext_conf_template.txt.
This includes:
provider keys and models
default engine selection
token/temperature values
basic auth settings
Caching
The extension registers cache nst3af_statistics in
ext_localconf.php.
Statistics service stores processed data in this cache to reduce repeated usage API calls.
Constraints
No native frontend plugin and no Fluid frontend output in this package.
Primary role is reusable service infrastructure.