This is the changelog for Authress SDK.
- Optimize JWKs fetching using the keyId
- [Breaking] Added type checking everywhere - This means most models have breaking changes.
- [Breaking] Converted properties to be consistent across all generators
- [Breaking] User ID is now explicitly required in the user permissions api methods to ensure that the correct user ID is being passed.
- [Breaking] Moved
authress_client.users
api toauthress_client.user_permissions
. - [Breaking] Moved
authress_client.records
api toauthress_client.access_records
. - [Breaking] Moved
authress_client.clients
api toauthress_client.service_clients
. - [Breaking] Renamed
AccessRecordStatement
model toStatement
inmodels.statement.py
. - [Breaking] Renamed
AccessRecordResource
model toResource
inmodels.resource.py
. - [Breaking] Renamed parameter
if_unmodified_since
in the Access Record API to beexpected_last_modified_time
. - Add missing
If-Unmodified-Since
support to theupdate_group
in theGroups
API. - Improve caching in
verify_token
- Support additionally pydantic v2 dependencies. Support for v1 will be removed in a future version.
- Prevent unnecessary extra trailing slashes in domain name.
- Add support for the
generateUserLoginUrl
from the ServiceClientTokenProvider. - [Breaking] Removed client_id from ClientAccessKey.py model, because it does not contain a clientId when fetching client data.
- Add support for
tokenConfiguration
at the tenant level.
- Add support for users and groups at the statement level of access records.
- Fix naming of some incorrectly named plural classes in the Access Record Models.
- Rename
HttpClient
toAuthressClient
and split out non-essential functionality to a separate HttpClient. - Add
limit
,filter
,cursor
toget_records
endpoint.
- Add the
TokenVerifier
- Add EdDSA support
- Add
Groups
toAccessRecords
- Explicitly set the algorithms in decoding the user token to identify the user sub