Enum BackendError 

pub enum BackendError {
    RateLimit {
        provider: String,
        model: String,
        retry_after_seconds: Option<u64>,
        body_preview: String,
    },
    Auth {
        provider: String,
        model: String,
        api_key_env: Option<String>,
        status: u16,
        body_preview: String,
    },
    ContextLength {
        provider: String,
        model: String,
        body_preview: String,
    },
    SafetyBreach {
        provider: String,
        model: String,
        finish_reason: String,
        body_preview: String,
    },
    ModelNotFound {
        provider: String,
        model: String,
        status: u16,
        body_preview: String,
    },
    Generic {
        provider: String,
        model: String,
        status: Option<u16>,
        message: String,
    },
}

Typed transport error from a native Rust backend.

Each variant carries enough context to render a useful message for the adopter without sacrificing programmatic dispatch — the message body repeats the provider name + model name + status code so log-only consumers don’t lose information.

Variants

RateLimit

Provider rate limit hit (HTTP 429); retries exhausted or unavailable.

Fields

provider: String

model: String

retry_after_seconds: Option<u64>

body_preview: String

Auth

Provider rejected the request as unauthenticated (HTTP 401 / 403).

Fields

provider: String

model: String

api_key_env: Option<String>

status: u16

body_preview: String

ContextLength

Compiled prompt exceeds the model’s context window (HTTP 400 with a context_length_exceeded / maximum context / too long shape).

Fields

provider: String

model: String

body_preview: String

SafetyBreach

Provider’s content filter blocked the request or response.

Fields

provider: String

model: String

finish_reason: String

body_preview: String

ModelNotFound

Provider does not recognise the requested model identifier.

Fields

provider: String

model: String

status: u16

body_preview: String

Generic

Unmapped HTTP error or transport-layer failure.

Fields

provider: String

model: String

status: Option<u16>

message: String

Implementations

impl BackendError

pub fn provider(&self) -> &str

Provider name reported by the error (always set).

pub fn model(&self) -> &str

Resolved model name reported by the error (always set; may be the provider’s default if the request omitted one).

pub fn kind(&self) -> BackendErrorKind

Translate into the legacy BackendErrorKind taxonomy so that existing infra in resilient_backend.rs / circuit_breaker.rs / retry_policy.rs continues to drive retry / CB decisions without changes during the Fase 24 transition (D6 — dual presence).

pub fn is_retryable(&self) -> bool

Whether the transport layer should retry this error before giving up. Consults Self::kind + the legacy is_retryable predicate so the policy stays in lockstep with resilient_backend.

pub fn category(&self) -> &'static str

Stable category label for logs / metrics. Preserved across variants so that operators can filter by error class without enum-matching.

Trait Implementations

impl Clone for BackendError

fn clone(&self) -> BackendError

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for BackendError

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Display for BackendError

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Error for BackendError

fn source(&self) -> Option<&(dyn Error + 'static)>

Returns the lower-level source of this error, if any.

fn description(&self) -> &str

👎Deprecated since 1.42.0:

use the Display impl or to_string()

fn cause(&self) -> Option<&dyn Error>

👎Deprecated since 1.33.0:

replaced by Error::source, which can support downcasting

fn provide<'a>(&'a self, request: &mut Request<'a>)

🔬This is a nightly-only experimental API. (error_generic_member_access)

Provides type-based access to context intended for error reports.