Error Handling

Handle API errors, authentication failures, token verification errors, and network issues in the Grantex Python SDK.

Overview

The Grantex Python SDK uses a structured exception hierarchy. All exceptions inherit from GrantexError, so you can catch all SDK errors with a single handler or handle specific error types individually.

Exception Hierarchy

GrantexError (base)
├── GrantexApiError (non-2xx HTTP response)
│   └── GrantexAuthError (401 or 403)
├── GrantexTokenError (JWT verification / decoding failure)
└── GrantexNetworkError (timeout, connection error)

Import

from grantex import (
    GrantexError,
    GrantexApiError,
    GrantexAuthError,
    GrantexTokenError,
    GrantexNetworkError,
)

GrantexError

The base exception for all Grantex SDK errors. Catch this to handle any SDK error:

from grantex import Grantex, GrantexError

with Grantex(api_key="gx_live_...") as client:
    try:
        agent = client.agents.get("agt_nonexistent")
    except GrantexError as e:
        print(f"Something went wrong: {e}")

GrantexApiError

Raised when the Grantex API returns a non-2xx HTTP response. Provides access to the HTTP status code, response body, and request ID.

from grantex import Grantex, GrantexApiError

with Grantex(api_key="gx_live_...") as client:
    try:
        agent = client.agents.get("agt_nonexistent")
    except GrantexApiError as e:
        print(f"Status code: {e.status_code}")
        print(f"Message: {e}")
        print(f"Response body: {e.body}")
        print(f"Request ID: {e.request_id}")

Attributes

Attribute	Type	Description
status_code	int	The HTTP status code (e.g. 404, 422, 500).
body	Any	The parsed response body (usually a dict) or raw text.
request_id	str | None	The X-Request-Id header value (if present).

The error message is extracted from the response body’s message or error field, falling back to "HTTP {status_code}".

GrantexAuthError

A subclass of GrantexApiError raised specifically for 401 (Unauthorized) and 403 (Forbidden) responses. Typically indicates an invalid or expired API key.

from grantex import Grantex, GrantexAuthError

with Grantex(api_key="gx_invalid_key") as client:
    try:
        agents = client.agents.list()
    except GrantexAuthError as e:
        print(f"Authentication failed ({e.status_code}): {e}")
        # Handle: refresh API key, prompt user, etc.

GrantexAuthError has the same attributes as GrantexApiError.

GrantexTokenError

Raised when JWT verification fails. This occurs when using verify_grant_token() for offline verification or when grants.verify() receives an active: false response (revoked, expired, or unknown token) from the Grantex API.

from grantex import verify_grant_token, VerifyGrantTokenOptions, GrantexTokenError

try:
    grant = verify_grant_token(
        "invalid.jwt.token",
        VerifyGrantTokenOptions(
            jwks_uri="https://api.grantex.dev/.well-known/jwks.json",
        ),
    )
except GrantexTokenError as e:
    print(f"Token error: {e}")

Common causes:

• Token uses an algorithm other than RS256
• Token signature is invalid
• Token is expired
• Required claims are missing (jti, sub, agt, dev, scp, iat, exp)
• Required scopes are not present
• JWKS endpoint is unreachable
• No matching RSA key found in the JWKS

GrantexNetworkError

Raised on network-level failures such as timeouts, DNS resolution errors, or connection refused. The original exception is available via the cause attribute.

from grantex import Grantex, GrantexNetworkError

with Grantex(api_key="gx_live_...", timeout=5.0) as client:
    try:
        agents = client.agents.list()
    except GrantexNetworkError as e:
        print(f"Network error: {e}")
        print(f"Original exception: {e.cause}")

Attributes

Attribute	Type	Description
cause	BaseException | None	The underlying httpx exception that caused the failure.

Best Practices

Catch Specific Errors First

Order your except clauses from most specific to least specific:

from grantex import (
    Grantex,
    GrantexAuthError,
    GrantexApiError,
    GrantexNetworkError,
    GrantexError,
)

with Grantex(api_key="gx_live_...") as client:
    try:
        agent = client.agents.get("agt_abc123")
    except GrantexAuthError as e:
        # 401/403 -- invalid or expired API key
        print(f"Auth failed: {e}")
    except GrantexApiError as e:
        # Other HTTP errors (404, 422, 500, etc.)
        print(f"API error {e.status_code}: {e}")
    except GrantexNetworkError as e:
        # Timeout, connection refused, DNS failure
        print(f"Network error: {e}")
    except GrantexError as e:
        # Catch-all for any other SDK error
        print(f"Unexpected error: {e}")

Retry on Transient Errors

Network errors and certain HTTP status codes (429, 502, 503, 504) are often transient and safe to retry:

import time
from grantex import Grantex, GrantexApiError, GrantexNetworkError

def get_agent_with_retry(client, agent_id, max_retries=3):
    for attempt in range(max_retries):
        try:
            return client.agents.get(agent_id)
        except GrantexNetworkError:
            if attempt == max_retries - 1:
                raise
            time.sleep(2 ** attempt)  # Exponential backoff
        except GrantexApiError as e:
            if e.status_code in (429, 502, 503, 504):
                if attempt == max_retries - 1:
                    raise
                time.sleep(2 ** attempt)
            else:
                raise  # Non-retryable API error

Use Request ID for Support

When reporting issues, include the request_id from GrantexApiError:

from grantex import Grantex, GrantexApiError

with Grantex(api_key="gx_live_...") as client:
    try:
        client.agents.delete("agt_abc123")
    except GrantexApiError as e:
        print(f"Error: {e}")
        if e.request_id:
            print(f"Request ID for support: {e.request_id}")