Skip to main content

​
Error Types

The Go SDK uses typed errors that implement the error interface. Use type assertions to handle specific error types.

​
APIError

Returned for any non-2xx HTTP response (except 401/403). 
type APIError struct {
    StatusCode int
    Body       json.RawMessage
    Code       string    // "BAD_REQUEST", "NOT_FOUND", etc.
    RequestID  string
    Message    string
}

​
AuthError

Returned for 401 (Unauthorized) and 403 (Forbidden) responses. Embeds *APIError. 
type AuthError struct {
    *APIError
}

​
TokenError

Returned from VerifyGrantToken() for token verification failures. 
type TokenError struct {
    Message string
    Cause   error    // Underlying error (implements Unwrap)
}

​
NetworkError

Returned for network-level failures (DNS, timeout, connection refused). 
type NetworkError struct {
    Message string
    Cause   error    // Underlying error (implements Unwrap)
}

​
Handling Pattern

agent, err := client.Agents.Get(ctx, "agent-id")
if err != nil {
    switch e := err.(type) {
    case *grantex.AuthError:
        // 401 or 403 — invalid or expired API key
        log.Printf("Auth failed (%d): %s", e.StatusCode, e.Message)
    case *grantex.APIError:
        // Other API errors (400, 404, 422, 500, etc.)
        log.Printf("API error %d [%s]: %s", e.StatusCode, e.Code, e.Message)
    case *grantex.NetworkError:
        // Network issues — retry or fail gracefully
        log.Printf("Network error: %s (cause: %v)", e.Message, e.Cause)
    default:
        log.Printf("Unexpected: %v", err)
    }
}

​
Unwrap Support

TokenError and NetworkError implement Unwrap() for use with errors.Is() and errors.As(): 
import "errors"

_, err := grantex.VerifyGrantToken(ctx, token, opts)
if err != nil {
    var tokenErr *grantex.TokenError
    if errors.As(err, &tokenErr) {
        fmt.Printf("Token error: %s\n", tokenErr.Message)
        if tokenErr.Cause != nil {
            fmt.Printf("Caused by: %v\n", tokenErr.Cause)
        }
    }
}