]> Defakto Security

arndts.ietf@gmail.com

Defakto Security

pieter@defakto.security

NIST

scott.rose@nist.gov

IBM

sthorger@ibm.com

Cisco Systems

ncamwing@cisco.com

sec Web Authorization Protocol workload identity credential exchange This specification profiles the Assertion Framework for OAuth 2.0 Client Authentication and Authorization Grants , the JWT Profile for OAuth 2.0 Client Authentication and Authorization Grants , and OAuth 2.0 Attestation-Based Client Authentication to enable the use of SPIFFE Verifiable Identity Documents (SVIDs) as client credentials in OAuth 2.0. It defines how OAuth clients with SPIFFE credentials can authenticate to OAuth authorization servers using their JWT-SVIDs, WIT-SVIDs, or X509-SVIDs without the need for client secrets. This approach enhances security by enabling seamless integration between SPIFFE-enabled workloads and OAuth authorization servers while eliminating the need to distribute and manage shared secrets such as static client secrets. About This Document Status information for this document may be found at . Discussion of this document takes place on the Web Authorization Protocol Working Group mailing list (), which is archived at . Subscribe at . Source for this draft and an issue tracker can be found at .

Introduction Traditional OAuth client authentication typically relies on client secrets or private key JWT authentication, both require an out of band distribution of secret material to the OAuth client. In modern cloud-native architectures where identity is managed by SPIFFE (Secure Production Identity Framework for Everyone), there is a need to provision additional secret material for OAuth clients when attested identifiers and credentials such as SVIDs are already available. This specification profiles the Assertion Framework for OAuth 2.0 Client Authentication and Authorization Grants to allow SPIFFE-enabled workloads to use their SPIFFE Verifiable Identity Documents (SVIDs), either X.509 certificates or JSON Web Tokens (JWT-SVID & WIT-SVID), as client credentials for OAuth 2.0 client authentication. JWT-SVIDs make use of a profiled version of the JWT Profile for OAuth 2.0 Client Authentication and Authorization Grants . WIT-SVIDs make use of the OAuth 2.0 Attestation-Based Client Authentication . This profile focuses on using SPIFFE credentials for OAuth client authentication. The SPIFFE profile for client authentication enables seamless integration between SPIFFE-based and OAuth-based systems, allowing applications to leverage both ecosystems without requiring additional credential management. It also enables a more secure authentication method by leveraging cryptographically verifiable credentials rather than shared secrets.

Conventions and Definitions The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 when, and only when, they appear in all capitals, as shown here.

Terminology This specification uses the terms defined in OAuth 2.0 , the Assertion Framework for OAuth 2.0 , the JWT profile of it , and the SPIFFE specifications. The following terms are used in this document: Trust Domain: As defined in SPIFFE; A trust domain represents a single trust root. All SVIDs issued within a trust domain are verifiable via the trust domain's keys. SPIFFE ID: A Uniform Resource Identifier (URI) that uniquely identifies a workload using the spiffe scheme. See for details. SVID: A SPIFFE Verifiable Identity Document. This document specifies the use of three types of SVIDs:

• X509-SVID: An X.509 certificate that contains a SPIFFE ID in the URI SAN extension. See for details.
• JWT-SVID: A JSON Web Token (JWT) that contains a SPIFFE ID in the sub claim. See for details.
• WIT-SVID: A WIMSE Workload Identity Token (WIT) that contains a SPIFFE ID in the sub claim. See for details.

SPIFFE Bundle: A collection of public keys and associated metadata that allow validation of SVIDs issued by a trust domain. SPIFFE Bundle Endpoint: A URL that serves a SPIFFE bundle for a trust domain.

OAuth Client Authentication Using SPIFFE This section describes how SPIFFE identity documents can be used for OAuth 2.0 client authentication, following the patterns established in and, for JWT-SVIDs, . OAuth 2.0 client authentication is used to authenticate the client to the authorization server when making requests to the token endpoint. When using SPIFFE for client authentication, the client presents its SVID (JWT-SVID, WIT-SVID, or X509-SVID) to prove its identity.

Client Identification The client_id request parameter is OPTIONAL for all client authentication methods defined in this specification. If the request contains a client_id parameter, the authorization server MUST verify that the SPIFFE ID of the presented SVID is authorized to authenticate as the client identified by the client_id:

• If the client_id is a URL pointing to a Client ID Metadata Document (), the SPIFFE ID MUST match the spiffe_id value in that document as described in .
• Otherwise, the SPIFFE ID MUST be associated with the client_id through the client's registration with the authorization server.

If the request does not contain a client_id parameter, the SPIFFE ID of the presented SVID is the client identifier. The authorization server MUST identify the client as the registered client whose client_id is equal to that SPIFFE ID. Clients identified by a Client ID Metadata Document MUST include the client_id parameter. In all other cases, the authorization server MUST reject the request with an invalid_client error as defined in Section 5.2 of . A client SHOULD include the client_id parameter whenever its client identifier differs from its SPIFFE ID.

Client Authentication with JWT-SVIDs JWT-SVID based authentication follows the JWT Profile for OAuth 2.0 Client Authentication , with specific adaptations for SPIFFE JWT-SVIDs. The requirements of continue to apply. To identify the assertion content as a JWT-SVID this specification establishes the following client assertion type as an OAuth URI according to : Based on the following request parameters are used to perform client authentication in the context of this specification:

• client_id: OPTIONAL. See .
• client_assertion_type: MUST be set to urn:ietf:params:oauth:client-assertion-type:jwt-spiffe.
• client_assertion: MUST be a single SPIFFE JWT-SVID.

To validate JWT-SVID client authentication requests the authorization server MUST:

1. Verify that the JWT is well-formed and contains all required claims (SPIFFE ID in sub, aud, and exp).
2. Verify that the JWT has not expired (check the exp claim).
3. Verify that the aud claim contains the issuer identifier of the authorization server as its sole value. See for details.
4. Verify the JWT signature using the signing keys of the trust domains according to .
5. Identify the client and verify that the SPIFFE ID in the sub claim is authorized to authenticate as that client according to .

JWT-SVID example The following example illustrates an authorization_code request to the token endpoint of an OAuth 2.0 authorization server leveraging a SPIFFE JWT-SVID to authenticate the client. The client is registered at the authorization server with client_id s6BhdRkqt3 and associated with the SPIFFE ID spiffe://example.org/my-oauth-client. For clarity, the JWT-SVID header and body decoded:

JWT-SVID example with Client ID Metadata Document The following example illustrates a client_credentials request to the token endpoint of an OAuth 2.0 authorization server leveraging a SPIFFE JWT-SVID to authenticate the client. For clarity, the JWT-SVID header and body decoded: The client_id points to a Client ID Metadata Document () with the following contents:

Client Authentication using X509-SVID X509-SVID based authentication uses mutual TLS as defined in OAuth 2.0 Mutual-TLS Client Authentication , with specific adaptations for SPIFFE X509-SVIDs. To authenticate using an X509-SVID, the client establishes a mutual TLS connection with the authorization server using its X509-SVID as the client certificate. The authorization server validates the client certificate as an X509-SVID and extracts the SPIFFE ID from the URI SAN. The client MUST authenticate the authorization server using standard HTTPS server certificate verification, with the trust anchors it uses for other HTTPS connections. The authorization server determines that X509-SVID client authentication applies from the client's registered authentication method (e.g., a token_endpoint_auth_method value of spiffe_x509), and not from the format of the client_id value. The client_id parameter is OPTIONAL. This deviates from Section 2 of , which requires the client_id parameter for mutual-TLS client authentication. Because the X509-SVID carries exactly one SPIFFE ID in its URI SAN, the client can be identified from the certificate without ambiguity. See . The authorization server validates the client certificate according to the following rules:

1. Validate the certificate as an X509-SVID according to .
2. Path validation MUST use only the trust anchors for the trust domain of the SPIFFE ID in the certificate's URI SAN, obtained as described in . Other trust anchors, including those of other trust domains or the system trust store, MUST NOT be used.
3. Identify the client and verify that the SPIFFE ID in the URI SAN is authorized to authenticate as that client according to .

Authorization servers MAY issue certificate-bound access tokens as described in Section 3 of to clients authenticated with an X509-SVID. Because X509-SVIDs are typically short-lived, such an access token cannot be used once the client's X509-SVID has been rotated, and the client needs to obtain a new access token.

X509-SVID Example The following request uses a refresh token to obtain a new access token. The client is s6BhdRkqt3, associated with the SPIFFE ID spiffe://example.org/my-oauth-client, and is authenticated by performing this request over a mutual TLS connection. For clarity, the presented X509-SVID client certificate to the server decoded via openssl x509 -text is:

Client Authentication with WIT-SVIDs A WIT-SVID is a WIMSE Workload Identity Token (WIT) as defined in that contains a SPIFFE ID in the sub claim . It uses OAuth 2.0 Attestation-Based Client Authentication for client authentication. A WIT-SVID as issued by a SPIFFE implementation binds a key pair held by the client via the cnf claim. The key pair bound to the WIT-SVID is used to generate an attestation proof during client authentication. The attestation proof is in the form of a "Client Attestation PoP JWT" as defined in that is issued by the client and signed with the private part of the key pair bound in the WIT-SVID. The WIT-SVID and the corresponding Client Attestation PoP JWT are sent together to the authorization server as a means of client authentication using the HTTP header-based syntax defined in Section 5.1 of . The client_id parameter is OPTIONAL. See .

Authorization Server Validation This specification is a profile of as described in Section 13 of that document. The authorization server determines that this profile applies when the typ JWT header parameter of the OAuth-Client-Attestation header is wit+jwt. To validate a WIT-SVID client authentication request, the authorization server MUST apply the following rules:

• The authorization server MUST accept wit+jwt as a valid value of the typ JWT header parameter in the OAuth-Client-Attestation header, in addition to oauth-client-attestation+jwt.
• The WIT-SVID MUST be validated as a Workload Identity Token according to Section 5.1.3 of .
• The WIT-SVID signature MUST be verified using the signing keys of the trust domain according to .
• The Client Attestation PoP JWT MUST be validated according to Section 7.2 of .
• The client MUST be identified and the SPIFFE ID in the sub claim MUST be verified according to . This replaces the client_id to sub matching rule of .

The authorization server advertises support for WIT-SVID client authentication using the spiffe_wit value as described in . The attest_jwt_client_auth value defined in indicates support for Client Attestation JWTs in general and does not by itself indicate support for WIT-SVIDs.

WIT-SVID Example The following example illustrates a token request using a WIT-SVID and a Client Attestation PoP JWT for client authentication. The client is registered at the authorization server with client_id s6BhdRkqt3 and associated with the SPIFFE ID spiffe://example.org/my-oauth-client. The WIT-SVID (OAuth-Client-Attestation) header and body decoded: The Client Attestation PoP JWT (OAuth-Client-Attestation-PoP) header and body decoded: The token endpoint request:

Interoperability In order to achieve interoperability between the authorization server and clients the authorization server MUST advertise what client authentication methods are supported. Authorization servers MUST support at least one of JWT-SVID, X509-SVID or WIT-SVID. The methods supported MUST be advertised in the authorization servers metadata by including spiffe_jwt, spiffe_wit and/or spiffe_x509 in the token_endpoint_auth_methods_supported list. Additionally, the same values should be included in revocation_endpoint_auth_methods_supported and introspection_endpoint_auth_methods_supported when applicable. Clients MUST support at least one of JWT-SVID, X509-SVID or WIT-SVID. To guarantee interoperability a client SHOULD support all. It is the responsibility of the client to select the authentication method supported by the authorization server and its deployment.

SPIFFE Trust Establishment and Client Registration This specification requires previously established trust between the OAuth 2.0 Authorization Server and the SPIFFE Trust Domain. This needs to happen out of band and is not in scope of this specification. However, the mechanism of key distribution is in scope and is described in . Similar to the trust establishment, corresponding OAuth clients need to be registered when using SPIFFE credentials for client authentication. OAuth client registration may leverage Client ID Metadata Documents , OAuth 2.0 Dynamic Client Registration or configure them out of band.

Client Registration Metadata This specification defines the following client metadata parameters for use in Client ID Metadata Documents :

spiffe_id

REQUIRED. The SPIFFE ID of the client, e.g. spiffe://example.org/my-oauth-client. The value MAY include a trailing /* character sequence (e.g. spiffe://example.org/workloads/*) indicating that a prefix match against the SPIFFE ID of the presented SVID is acceptable rather than requiring an exact match.

spiffe_bundle_endpoint

OPTIONAL. The URL of the SPIFFE Bundle Endpoint for the client's trust domain, which the authorization server can use to retrieve the signing keys for validating the client's SVIDs. If not provided, the authorization server MUST obtain the signing keys for the client's trust domain through some other established mechanism, such as a pre-configured SPIFFE bundle. Bundles obtained from this endpoint are subject to . See also .

If the spiffe_id value uses a wildcard, it MUST end with the two-character sequence "/*". In this case, the authorization server MUST perform a path-segment prefix match: the SPIFFE ID of the presented SVID MUST begin with the spiffe_id value excluding the trailing "*" (i.e., up to and including the final "/"), and this prefix MUST correspond to complete path segment(s) of the SPIFFE ID (for example, spiffe://example.org/client/* matches spiffe://example.org/client/123 but does not match spiffe://example.org/client123). Otherwise, the SPIFFE ID of the presented SVID MUST be an exact match of the spiffe_id value. A client indicates the SPIFFE-based client authentication methods it supports in its client metadata, whether registered using Dynamic Client Registration or published in a Client ID Metadata Document , using the values defined in . A client that supports more than one method lists them in the token_endpoint_auth_methods_supported parameter defined in Section 2 of . A client SHOULD also include the token_endpoint_auth_method parameter with its preferred method, so that authorization servers that do not support token_endpoint_auth_methods_supported do not apply the default of client_secret_basic.

SPIFFE Key Distribution and Validation This section describes how an authorization server verifies the signature of SVIDs. It recommends two SPIFFE-native approaches. The 2 approaches can be used in conjunction, for instance: Trust bundles in general MUST be keyed by the trust domain identifier to prevent mix-up between trust domains and their corresponding bundles. Bundles obtained from client metadata are additionally keyed by the client_id, as described in .

Bundles from Client Metadata When an authorization server obtains a bundle from a client's spiffe_bundle_endpoint, it MUST:

• associate the bundle with both the client_id and the trust domain of the client's spiffe_id, and use it only to validate SVIDs presented to authenticate as that client;
• not add the bundle's keys to any bundle it holds for that trust domain, nor use them to validate SVIDs presented for any other client.

If the authorization server has a bundle for the trust domain obtained as described in , it SHOULD use that bundle and ignore spiffe_bundle_endpoint.

SPIFFE Bundle Endpoint The SPIFFE Bundle Endpoint as defined in Section 4 of exposes the signing keys for SVIDs over HTTP via a JSON Web Key Set according to . SPIFFE Federation defines two endpoint profiles for serving and consuming a bundle endpoint. For interoperability, the authorization server MUST use the https_web profile (Section 5.2.1 of ). The bundle endpoint is served over HTTPS, and the authorization server verifies the endpoint's server certificate using standard HTTPS server certificate verification against its locally configured trust anchors. This specification does not use the https_spiffe profile. The authorization server SHOULD manage and refresh trust bundles according to Section 4.1 and 4.2 of . The authorization server MUST associate each bundle endpoint with exactly one trust domain, and MUST use keys obtained from that endpoint only to validate SVIDs. At the time of writing, the SPIFFE bundle endpoint cannot be derived from an SVID or SPIFFE-ID and MUST be configured out of band. The security considerations that apply when configuring it using spiffe_bundle_endpoint of Client ID Metadata Documents are described in .

Example The following examples showcase how the Authorization Server can perform key discovery for the trust domain example.org. Important to note is the difference between example.org trust domain and example.com location for the SPIFFE Bundle Endpoint. This highlights the importance of explicit configuration and demonstrates why the SPIFFE Bundle Endpoint cannot be derived or discovered from SVIDs without explicit configuration. Example configuration at the OAuth Authorization Server in the JSON format

• Note difference between example.org and example.com

Example SPIFFE Bundle Endpoint request, response:

• The use parameter in the JSON Web Key indicates the credential format the key is intended for. Multiple keys of the same use can be present.

The X509-SVID signing certificate (.keys[0].x5c[0] from response above) in text form:

Alternative methods to avoid The following key distribution mechanisms are alternatives and SHOULD be avoided for interoperability reasons.

SPIFFE Workload API The SPIFFE Workload API allows workloads to retrieve a trust bundle. It requires the authorization server to be part of a SPIFFE trust domain and be considered a workload within it. The SPIFFE Workload API is designed so that the workload receives trust bundle updates as they occur and does not need to poll for them, which reduces the time to distribute them. In addition to the trust bundle of the trust domain the workload resides in, the SPIFFE Workload API also allows retrieval of trust bundles from federated trust domains. This approach is NOT RECOMMENDED for OAuth SPIFFE Client Authentication for the following reasons:

• OAuth Authorization Server needs to be a workload within a SPIFFE trust domain, which is a significant limitation for deployment scenarios.
• Federated trust domain bundles create ambiguity about how they are handled. When distributed via the SPIFFE Workload API the trust relationship and points where they are established become ambiguous.

Manual configuration In small, static environments the authorization server MAY be configured with the SPIFFE bundles manually. This approach requires human interaction to set up, rotate and manage keying material and is thus generally NOT RECOMMENDED.

Using the system trust store X509-SVIDs MUST NOT be validated using the system trust store. The SPIFFE ID carried in the URI SAN is usually not a verifiable attribute in the broader X.509 ecosystem. Using the system trust store as trust anchor would allow ANY certificate authority in it to issue a trusted X509-SVID for ANY SPIFFE ID. In comparison: using SPIFFE-native validation methods restricts the signing of SPIFFE-IDs to the corresponding trust domain signing keys.

Using the JWT-SVID or WIT-SVID iss claim JSON Web Token-based credentials carrying iss claims could technically be validated by retrieving the signing keys via OpenID Connect Discovery or OAuth 2.0 Authorization Server Metadata. This approach only applies for the JWT-SVID and WIT-SVID and only works when the iss claim is present, which is optional. Because of its narrow scope and interoperability considerations, this approach is not a general alternative to the SPIFFE Bundle Endpoint. Implementations SHOULD use the mechanisms defined in this specification when available. However, when those mechanisms are not supported by a peer deployment, implementations MAY use iss-based discovery and key retrieval for JWT-SVID or WIT-SVID validation as a compatibility mechanism, subject to local policy, appropriate trust configuration and risk mitigations described in .

Implementation Status Note to RFC Editor: please remove this section, as well as the reference to RFC 7942, before publication. This section records the status of known implementations of the protocol defined by this specification at the time of posting of this Internet-Draft, and is based on a proposal described in . The description of implementations in this section is intended to assist the IETF in its decision processes in progressing drafts to RFCs. Please note that the listing of any individual implementation here does not imply endorsement by the IETF. Furthermore, no effort has been spent to verify the information presented here that was supplied by IETF contributors. This is not intended as, and must not be construed to be, a catalog of available implementations or their features. Readers are advised to note that other implementations may exist. According to RFC 7942, "this will allow reviewers and working groups to assign due consideration to documents that have the benefit of running code, which may serve as evidence of valuable experimentation and feedback that have made the implemented protocols more mature. It is up to the individual working groups to use this information as they see fit". Keycloak

• Organization: CNCF / IBM
• Maturity: preview
• Coverage: JWT-SVID client authentication using SPIFFE Trust Bundle Endpoint
• Contact: Keycloak community & maintainers

Security Considerations Client authentication using JWT-SVIDs has the same security considerations as described in , and . Client authentication using X509-SVIDs has the same security considerations as described in . The validation rules in protect against an OAuth2 token being issued (or being issued incorrectly) to a client that did not present an appropriate X509-SVID. Client authentication using WIT-SVIDs has the same security considerations as described in and . The issues described in include the risk that an authorization server may have the incorrect trust stores configured to validate the client SVID. This could result in an incorrectly issued token to an attacker if the attacker is able to obtain a certificate that can be validated by one of the misconfigured trust anchors in the trust store.

Replay of JWT-SVIDs permits but does not require a jti claim, and JWT-SVID validators are not required to track jti uniqueness. SPIFFE implementations may also return the same cached JWT-SVID to a workload for multiple requests. Authorization servers therefore cannot in general rely on jti-based replay detection as described in . An authorization server may use the jti claim for replay detection where it is present and the deployment does not reuse JWT-SVIDs across requests. An attacker that obtains a JWT-SVID can use it for client authentication until it expires. The audience restriction in step 3 of limits its use to this authorization server. Deployments requiring replay resistance SHOULD either use JWT-SVIDs that carry a unique jti claim and are not reused, with the authorization server rejecting a repeated jti, or use X509-SVID or WIT-SVID client authentication, which require proof of possession of a private key.

Trust Bundles from Client ID Metadata Documents When an authorization server accepts spiffe_bundle_endpoint from a Client ID Metadata Document, the party controlling the client_id URL also selects the keys used to validate the client's SVIDs. Successful authentication then shows that the client holds a key vouched for by that party, not that the trust domain named in spiffe_id issued the SVID. The trust model rests on control of the client_id host and the integrity of HTTPS retrieval, as described in the Security Considerations of . In return, trust bundles can be obtained without configuring a bundle endpoint for each trust domain in advance. Deployments that do not accept this trust model MUST ignore spiffe_bundle_endpoint values in Client ID Metadata Documents and SHOULD obtain bundles as described in . If keys from a document were used beyond the client it describes, an attacker controlling a single document could introduce a signing key of its choice, issue SVIDs for any SPIFFE ID in that trust domain, and authenticate as any other client associated with that trust domain. The rules in prevent this. They extend the binding between trust domain and bundle described in Section 7.3 of to include the client_id. Client scoping does not prevent a document from claiming a trust domain its author does not operate. Trust domain names are chosen by their operators and are not bound to DNS (), so a document hosted at https://attacker.example/ can declare a spiffe_id of spiffe://bank.example/payments together with its own bundle. The attacker can then authenticate only as its own client, but the SPIFFE ID it presents was not issued by the bank.example trust domain. Unless the bundle was obtained as described in , or the trust domain is bound to the host of the client_id URL, the authorization server MUST NOT treat that SPIFFE ID as issued by the trust domain, for example in authorization policy, access token claims or introspection responses. Authorization servers that accept bundles from documents SHOULD apply one or more of the following mitigations:

• require the trust domain of spiffe_id to equal, or be a subdomain of, the host of the client_id URL;
• require spiffe_id to be an exact SPIFFE ID rather than a wildcard;
• restrict which client_id URLs are accepted using domain allowlists or pre-registration as described in ;
• issue short-lived or sender-constrained access tokens to such clients;
• treat a change in spiffe_id, spiffe_bundle_endpoint, or the keys it serves as a change in client keys, as described in .

JWT-SVID and WIT-SVID iss claim As described in , iss-based key discovery is not a general alternative to the SPIFFE Bundle Endpoint and SHOULD be avoided. However, if it is used, the authorization server MUST NOT perform issuer-based discovery solely based on the iss value from the token unless the iss value is already trusted by the authorization server, and that iss value is explicitly associated with the configured SPIFFE Trust Domain being validated. In addition implementations MUST NOT assume that keys advertised via OpenID Connect Discovery or OAuth 2.0 Authorization Server Metadata are dedicated to JWT-SVID issuance. The same provider infrastructure may issue multiple JWT types (e.g., OAuth/OIDC tokens, JWT-SVIDs, WIT-SVIDs). A token MUST NOT be accepted as a JWT-SVID or WIT-SVID solely because it is a JWT, its signature validates under keys discovered from iss, and its sub resembles a SPIFFE ID. Doing so can enable token confusion (e.g., presenting an OAuth/OIDC token issued for another purpose as a JWT-SVID). Implementations MUST validate SVIDs according to SVID-specific requirements and MUST use a trust anchor or key source explicitly bound to the configured SPIFFE Trust Domain, rather than generic issuer-based discovery from untrusted token input.

IANA Considerations This document requests a new entry to be added to the "OAuth URI" registry found at https://www.iana.org/assignments/oauth-parameters/oauth-parameters.xhtml#uri. The registration process is defined in . This document requests the following entry to be added to the registry:

• URN: urn:ietf:params:oauth:client-assertion-type:jwt-spiffe
• Common Name: SPIFFE JWT-SVID Profile for OAuth 2.0 Client Authentication
• Change Controller: IETF
• Reference: This Document

OAuth Dynamic Client Registration Metadata This document requests the following entries to be added to the "OAuth Dynamic Client Registration Metadata" registry defined in :

• Client Metadata Name: spiffe_id
• Client Metadata Description: SPIFFE ID of the client
• Change Controller: IETF
• Reference: This Document
• Client Metadata Name: spiffe_bundle_endpoint
• Client Metadata Description: URL of the SPIFFE Bundle Endpoint for the client's trust domain
• Change Controller: IETF
• Reference: This Document

OAuth Token Endpoint Authentication Methods This document requests the following entries in the "OAuth Token Endpoint Authentication Methods" registry:

• Token Endpoint Authentication Method Name: spiffe_jwt
• Change Controller: IETF
• Reference: of this document
• Token Endpoint Authentication Method Name: spiffe_x509
• Change Controller: IETF
• Reference: of this document
• Token Endpoint Authentication Method Name: spiffe_wit
• Change Controller: IETF
• Reference: of this document

References Normative References The OAuth 2.0 authorization framework enables a third-party application to obtain limited access to an HTTP service, either on behalf of a resource owner by orchestrating an approval interaction between the resource owner and the HTTP service, or by allowing the third-party application to obtain access on its own behalf. This specification replaces and obsoletes the OAuth 1.0 protocol described in RFC 5849. [STANDARDS-TRACK] This document establishes an IETF URN Sub-namespace for use with OAuth-related specifications. This document is not an Internet Standards Track specification; it is published for informational purposes. A JSON Web Key (JWK) is a JavaScript Object Notation (JSON) data structure that represents a cryptographic key. This specification also defines a JWK Set JSON data structure that represents a set of JWKs. Cryptographic algorithms and identifiers for use with this specification are described in the separate JSON Web Algorithms (JWA) specification and IANA registries established by that specification. This specification provides a framework for the use of assertions with OAuth 2.0 in the form of a new client authentication mechanism and a new authorization grant type. Mechanisms are specified for transporting assertions during interactions with a token endpoint; general processing rules are also specified. The intent of this specification is to provide a common framework for OAuth 2.0 to interwork with other identity systems using assertions and to provide alternative client authentication mechanisms. Note that this specification only defines abstract message flows and processing rules. In order to be implementable, companion specifications are necessary to provide the corresponding concrete instantiations. This specification defines the use of a JSON Web Token (JWT) Bearer Token as a means for requesting an OAuth 2.0 access token as well as for client authentication. This specification defines mechanisms for dynamically registering OAuth 2.0 clients with authorization servers. Registration requests send a set of desired client metadata values to the authorization server. The resulting registration responses return a client identifier to use at the authorization server and the client metadata values registered for the client. The client can then use this registration information to communicate with the authorization server using the OAuth 2.0 protocol. This specification also defines a set of common client metadata fields and values for clients to use during registration. This document describes OAuth client authentication and certificate-bound access and refresh tokens using mutual Transport Layer Security (TLS) authentication with X.509 certificates. OAuth clients are provided a mechanism for authentication to the authorization server using mutual TLS, based on either self-signed certificates or public key infrastructure (PKI). OAuth authorization servers are provided a mechanism for binding access tokens to a client's mutual-TLS certificate, and OAuth protected resources are provided a method for ensuring that such an access token presented to it was issued to the client presenting the token. This specification defines a metadata format that an OAuth 2.0 client can use to obtain the information needed to interact with an OAuth 2.0 authorization server, including its endpoint locations and authorization server capabilities. Okta This specification defines a mechanism through which an OAuth client can identify itself to authorization servers, without prior dynamic client registration or other existing registration. This is through the usage of a URL as a client_id in an OAuth flow, where the URL refers to a document containing the necessary client metadata, enabling the authorization server to fetch the metadata about the client as needed. MATTR Bundesdruckerei SPRIND This specification defines an extension to the OAuth 2.0 protocol (RFC 6749) that enables a client instance to include a key-bound attestation when interacting with an Authorization Server or Resource Server. This mechanism allows a client instance to prove its authenticity, as verified by a client attester, without revealing its target audience to that attester. It may also serve as a mechanism for client authentication as per OAuth 2.0. Ping Identity Palo Alto Networks Defakto Security Independent Zscaler The WIMSE architecture defines authentication and authorization for software workloads in a variety of runtime environments, from the most basic ones up to complex multi-service, multi-cloud, multi- tenant deployments. This document defines the credentials that workloads use to represent their identity. They can be used in various protocols to authenticate workloads to each other. To use these credentials, workloads must provide proof of possession of the associated private key material, which is covered in other documents. This document focuses on the credentials alone, independent of the proof-of- possession mechanism. n.d. n.d. n.d. n.d. n.d. n.d. In many standards track documents several words are used to signify the requirements in the specification. These words are often capitalized. This document defines these words as they should be interpreted in IETF documents. This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements. RFC 2119 specifies common key words that may be used in protocol specifications. This document aims to reduce the ambiguity by clarifying that only UPPERCASE usage of the key words have the defined special meanings. Self-Issued Consulting Ping Identity Disney Okta This document updates RFC7521, RFC7522, RFC7523 and RFC9126 with respect to the treatment of audience values in OAuth 2.0 Client Assertion Authentication and Assertion-based Authorization Grants to address a security vulnerability identified in the previous requirements for those audience values in multiple OAuth 2.0 specifications. This document describes a simple process that allows authors of Internet-Drafts to record the status of known implementations by including an Implementation Status section. This will allow reviewers and working groups to assign due consideration to documents that have the benefit of running code, which may serve as evidence of valuable experimentation and feedback that have made the implemented protocols more mature. This process is not mandatory. Authors of Internet-Drafts are encouraged to consider using the process for their documents, and working groups are invited to think about applying the process to all of their protocol specifications. This document obsoletes RFC 6982, advancing it to a Best Current Practice. Informative References

Document History RFC Editor: please remove before publication.

draft-ietf-oauth-spiffe-client-auth-03

• Make WIT and attestation based client authentication references normative
• Make the client_id parameter optional for all client authentication methods and define unified client identification rules.
• Regenerate the examples for consistency (see https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/50)
• Describe the trust model and mitigations for obtaining trust bundles from Client ID Metadata Documents (see https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/55)
• Fixed Nits (see https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/67)
• Reference X509-SVID validation rules instead of restating them (https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/52)
• Describe how clients indicate one or more SPIFFE client authentication method (https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/49)
• Clarify that the authorization server selects X509-SVID client authentication from the client's registered authentication method (see https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/31)
• Align terminology with SPIFFE (see https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/45)
• Register spiffe_jwt, spiffe_x509 and spiffe_wit in the OAuth Token Endpoint Authentication Methods registry (see https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/48)
• Clarify Authorization Server Authentication when using mutual TLS.
• Clarify use of https_web and https_spiffe profiles (https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/22 and https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/27)
• Provide guidance on jti parameter (see https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/20)
• Added guidance on credential rotation and access token binding (https://github.com/oauth-wg/oauth-spiffe-client-authentication/issues/11)

draft-ietf-oauth-spiffe-client-auth-02

• Add Nancy Cam-Winget as co-author.
• Editorial updates to clarify WIT-SVID support.

draft-ietf-oauth-spiffe-client-auth-01

• Add mechanism to use Client ID Metadata Document for SPIFFE client metadata discovery.
• Add WIT-SVID variant using Workload Identity Tokens and Attestation-Based Client Authentication.
• Add interoperability section.
• Add more guidance and security considerations when using the iss claim.
• Add Stian Thorgersen as co-author.

draft-ietf-oauth-spiffe-client-auth-00

• Document name update to reflect adoption into the OAuth working group

draft-schwenkschuster-oauth-spiffe-client-auth-01

• Rephrase introduction to make the focus on client authentication more clear.
• Add implementation section.
• Add audience restrictions from RFC7523bis adopted WG document.
• Add security and IANA considerations section.
• Add Scott Rose as co-author.

draft-schwenkschuster-oauth-spiffe-client-auth-00

• Initial document

Acknowledgments The authors thank Joe DeCock, Rajvardhan Deshmukh, Takahiko Kawasaki, Marcel Levy, Aaron Parecki and Dag Sneeggen for their reviews, issues and pull requests.