Documentation
¶
Overview ¶
Package jev implements a decision.provider capability backed by a Jev-style "System One" decision API: state plus typed questions in, typed answers with calibrated probabilities out.
The endpoint is configurable because the same wire shape is served from several places: a local LM Studio instance hosting an open Jev-compatible model (the default — LM Studio exposes POST /v1/systemone for loaded decision models), TypeSafe AI's hosted API, or a gateway. The Vercel AI Gateway's /v1/evaluate names the yes/no type "boolean" and returns "probability" rather than "noul"; the `dialect` key selects that variant.
Jev is not an LLM and does not ride llm.request: it answers decision.ask events, filling the pointer payload in place like the search providers do.
Index ¶
- func New() engine.Plugin
- type Plugin
- func (p *Plugin) Capabilities() []engine.Capability
- func (p *Plugin) ConfigSchema() []byte
- func (p *Plugin) Dependencies() []string
- func (p *Plugin) Emissions() []string
- func (p *Plugin) ID() string
- func (p *Plugin) Init(ctx engine.PluginContext) error
- func (p *Plugin) Name() string
- func (p *Plugin) Ready() error
- func (p *Plugin) Requires() []engine.Requirement
- func (p *Plugin) Shutdown(_ context.Context) error
- func (p *Plugin) Subscriptions() []engine.EventSubscription
- func (p *Plugin) Version() string
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
Types ¶
type Plugin ¶
type Plugin struct {
// contains filtered or unexported fields
}
Plugin advertises decision.provider and answers decision.ask events.
func (*Plugin) Capabilities ¶
func (p *Plugin) Capabilities() []engine.Capability
func (*Plugin) ConfigSchema ¶
ConfigSchema implements engine.ConfigSchemaProvider.
func (*Plugin) Dependencies ¶
func (*Plugin) Requires ¶
func (p *Plugin) Requires() []engine.Requirement
func (*Plugin) Subscriptions ¶
func (p *Plugin) Subscriptions() []engine.EventSubscription