Struct ProviderLlm 

pub struct ProviderLlm {
    pub port: Arc<dyn LlmPort>,
    pub model_id: String,
    pub mechanism: ExtractMechanism,
    pub model_digest: Option<String>,
    pub tokenizer: Option<Arc<dyn TokenizerPort>>,
    pub source: ResolutionSource,
}

A resolved LLM provider: the port plus everything ExtractorConfig and Brain need to reflect it (model id, mechanism, weights digest, exact tokenizer when the provider ships one, plus the resolution source for provenance).

Fields

port: Arc<dyn LlmPort>

model_id: String

mechanism: ExtractMechanism

model_digest: Option<String>

blake3 hex digest of the weights, when the provider is a local artifact (§9.5 — weight changes must invalidate the extraction cache).

tokenizer: Option<Arc<dyn TokenizerPort>>

source: ResolutionSource

Where the resolution ladder picked this provider. Task 5 threads profile identity / model digest into ExtractorId provenance from this field; the CLI does not edit ExtractorConfig directly here.

Implementations

impl ProviderLlm

pub fn profile_id(&self) -> Option<String>

Foundation profile id when the provider came from a profile resolution; None for the legacy compat env / explicit override / local GGUF paths. Consumed by Task 5 to fold the binding into ExtractorConfig::provider_profile_id and invalidate cached summaries when the role changes (§13).