The OAuth 2.0 Authorization Protocol: Bearer TokensMicrosoftmbj@microsoft.comhttp://self-issued.info/independentdick.hardt@gmail.comhttp://dickhardt.org/Facebookdr@fb.comhttp://www.davidrecordon.com/
Security
OAuth Working Group
This specification describes how to use bearer tokens in HTTP
requests to access OAuth 2.0 protected resources. Any party
in possession of a bearer token (a "bearer") can use it to get
access to the associated resources (without demonstrating possession
of a cryptographic key). To prevent misuse, bearer tokens
need to be protected from disclosure in storage and in transport.
OAuth enables clients to access protected resources by
obtaining an access token, which is defined in
OAuth 2.0 Authorization
as "a string representing an access
authorization issued to the client", rather than using the
resource owner's credentials directly.
Tokens are issued to clients by an authorization server with the approval of
the resource owner. The client uses the access token to access the protected resources
hosted by the resource server. This specification describes how to make protected resource
requests when the OAuth access token is a bearer token.
This specification defines the use of bearer tokens over
HTTP/1.1
using
TLS to access protected resources.
TLS is mandatory to implement
and use with this specification; other specifications may
extend this specification for use with other transport
protocols.
While designed for use with access tokens resulting from
OAuth 2.0 Authorization
flows to access OAuth protected resources, this
specification actually defines a general HTTP authorization
method that can be used with bearer tokens from any source
to access any resources protected by those bearer tokens.
The key words 'MUST', 'MUST NOT', 'REQUIRED', 'SHALL', 'SHALL NOT', 'SHOULD', 'SHOULD
NOT', 'RECOMMENDED', 'MAY', and 'OPTIONAL' in this document are to be interpreted as
described in
Key words for use in RFCs to Indicate Requirement Levels .
This document uses the Augmented Backus-Naur Form (ABNF)
notation of
HTTP/1.1, Part 1 ,
which is based upon the
Augmented Backus-Naur Form (ABNF)
notation. Additionally, the following rules are included from
HTTP/1.1, Part 7 :
auth-param, auth-scheme, and b64token; and from
Uniform Resource Identifier (URI) :
URI-Reference.
Unless otherwise noted, all the protocol parameter names and values are case sensitive.
A security token with the property that any party in
possession of the token (a "bearer") can use the token
in any way that any other party in possession of it can.
Using a bearer token does not require a bearer to prove
possession of cryptographic key material
(proof-of-possession).
All other terms are as defined in
OAuth 2.0 Authorization .
OAuth provides a method for clients to access a protected resource on behalf of a
resource owner. In the general case,
before a client can access a protected resource, it must first obtain
an authorization grant from the resource owner and then exchange the authorization grant for
an access token.
The access token represents the grant's scope, duration, and
other attributes granted by the authorization grant. The
client accesses the protected resource by presenting the
access token to the resource server.
In some cases, a client can directly present its own
credentials to an authorization server to obtain an access
token without having to first obtain an authorization grant from a
resource owner.
The access token provides an abstraction, replacing different authorization
constructs (e.g. username and password, assertion) for a single token understood by the
resource server. This abstraction enables issuing access tokens valid for a short time
period, as well as removing the resource server's need to understand a wide range of
authentication schemes.
The abstract flow illustrated in describes the overall
OAuth 2.0 protocol architecture. The following steps are specified within this
document:
E) The client makes a protected resource request to the resource server by presenting
the access token.
F) The resource server validates the access token, and if valid, serves the request.
This section defines three
methods of sending bearer access tokens in resource requests
to resource servers. Clients MUST NOT use more than one
method to transmit the token in each request.
When sending the access token in the Authorization request header field
defined by
HTTP/1.1, Part 7 ,
the
client uses the Bearer
authentication scheme to transmit the access token.
The Authorization header field uses the framework defined by
HTTP/1.1, Part 7
as follows:
The b64token syntax was chosen over the alternative
#auth-param syntax also defined by
HTTP/1.1, Part 7
both for simplicity
and for compatibility with existing implementations.
If additional parameters are needed in the future, a
different scheme would need to be defined.
Clients SHOULD make authenticated requests with a bearer
token using the Authorization
request header field with the Bearer HTTP authorization scheme.
Resource servers MUST support this method.
When sending the access token in the HTTP request
entity-body, the client adds the access token to the request
body using the access_token
parameter. The client MUST NOT use this method unless
all of the following conditions are met:
The HTTP request entity-header includes the Content-Type
header field set to application/x-www-form-urlencoded.
The entity-body follows the encoding requirements of the
application/x-www-form-urlencoded content-type as
defined by
HTML 4.01 .
The HTTP request entity-body is single-part.
The content to be encoded in the entity-body MUST
consist entirely of ASCII characters.
The HTTP request method is one for which the request
body has defined semantics. In particular,
this means that the GET
method MUST NOT be used.
The entity-body MAY include other request-specific
parameters, in which case, the access_token parameter MUST be properly
separated from the request-specific parameters using & character(s) (ASCII code 38).
The application/x-www-form-urlencoded
method SHOULD NOT be used except in application contexts
where participating browsers do not have access to the
Authorization request header
field. Resource servers MAY support this method.
When sending the access token in the HTTP request URI, the client adds the access
token to the request URI query component as defined by
Uniform Resource Identifier (URI)
using
the access_token parameter.
The HTTP request URI query can include other
request-specific parameters, in which case, the access_token parameter MUST be properly
separated from the request-specific parameters using & character(s) (ASCII code 38).
Because of the security weaknesses associated with the URI
method (see ), including the high
likelihood that the URL containing the access token will be
logged, it SHOULD NOT be used unless it is impossible to
transport the access token in the Authorization request header field or
the HTTP request entity-body. Resource servers MAY support
this method.
If the protected resource request does not include
authentication credentials or does not contain an access
token that enables access to the protected resource,
the resource server MUST include the HTTP WWW-Authenticate response header field;
it MAY include it in response to other conditions as well.
The WWW-Authenticate header
field uses the framework defined by
HTTP/1.1, Part 7 .
All challenges defined by this specification MUST use the
auth-scheme value Bearer. This
scheme MUST be followed by one or more auth-param values. The
auth-param attributes used or defined by this specification
are as follows. Other auth-param attributes MAY be used as
well.
A realm attribute MAY be included
to indicate the scope of protection in the manner described in
HTTP/1.1, Part 7 .
The realm attribute MUST NOT appear more than once.
The scope attribute is a space-delimited list of scope values
indicating the required scope of the access token for accessing the requested resource.
In some cases, the scope value
will be used when requesting a new access token with
sufficient scope of access to utilize the protected resource.
Use of the scope attribute is OPTIONAL.
The scope attribute MUST NOT appear more than once.
The scope value is intended for
programmatic use and is not meant to be displayed to
end users.
If the protected resource request included an access token and failed authentication, the
resource server SHOULD include the error attribute to provide
the client with the reason why the access request was declined. The parameter value is
described in .
In addition, the resource server MAY include the error_description attribute to provide
developers a human-readable explanation that is not meant
to be displayed to end users.
It also MAY include
the error_uri attribute with
an absolute URI identifying a human-readable web page explaining the error.
The error, error_description, and
error_uri attributes MUST NOT appear more than once.
Values for the scope attribute MUST NOT include
characters outside the set %x21 / %x23-5B / %x5D-7E
for representing scope values and %x20 for delimiters between scope values.
Values for the error and error_description attributes MUST NOT include
characters outside the set %x20-21 / %x23-5B / %x5D-7E.
Values for the error_uri attribute
MUST conform to the URI-Reference syntax, and thus MUST NOT include
characters outside the set %x21 / %x23-5B / %x5D-7E.
When a request fails, the resource server responds using the appropriate HTTP status
code (typically, 400, 401, 403, or 405),
and includes one of the following error codes in
the response:
The request is missing a required parameter, includes an unsupported parameter or
parameter value, repeats the same parameter, uses more than one method for
including an access token, or is otherwise malformed. The resource server SHOULD
respond with the HTTP 400 (Bad Request) status code.
The access token provided is expired, revoked, malformed, or invalid for other
reasons. The resource SHOULD respond with the HTTP 401 (Unauthorized) status
code. The client MAY request a new access token and retry the protected resource
request.
The request requires higher privileges than provided by the access token. The
resource server SHOULD respond with the HTTP 403 (Forbidden) status code and MAY
include the scope attribute with the scope necessary to
access the protected resource.
If the request lacks any authentication information (i.e. the client was unaware
authentication is necessary or attempted using an unsupported authentication method),
the resource server SHOULD NOT include an error code or other error information.
This section describes the relevant security threats regarding
token handling when using bearer tokens and describes how to
mitigate these threats.
The following list presents several common threats against
protocols utilizing some form of tokens. This list of
threats is based on
NIST Special Publication 800-63 .
Since this document builds on the
OAuth 2.0 specification, we exclude a discussion of threats
that are described there or in related documents.
An attacker may generate a bogus token or modify the
token contents (such as the authentication or attribute
statements) of an existing token, causing the resource
server to grant inappropriate access to the client.
For example, an attacker may modify the token to extend
the validity period; a malicious client may modify the
assertion to gain access to information that they
should not be able to view.
Tokens may contain authentication and attribute
statements that include sensitive information.
An attacker uses a token generated for consumption by
one resource server to gain access to a different
resource server that mistakenly believes the token to be
for it.
An attacker attempts to use a token that has already
been used with that resource server in the past.
A large range of threats can be mitigated by protecting the
contents of the token by using a digital signature or a
Message Authentication Code (MAC).
Alternatively, a bearer token can contain a reference to
authorization information, rather than encoding the
information directly. Such references MUST be infeasible for
an attacker to guess; using a reference may require an extra
interaction between a server and the token issuer to resolve
the reference to the authorization information.
The mechanics of such an interaction are not defined by this
specification.
This document does not specify the encoding or the contents
of the token; hence detailed recommendations about the means
of guaranteeing token integrity protection are outside the
scope of this document. The token integrity protection MUST
be sufficient to prevent the token from being modified.
To deal with token redirect, it is important for the
authorization server to include the identity of the intended
recipients (the audience), typically a single resource
server (or a list of resource servers), in the token.
Restricting the use of the token to a specific scope is also
RECOMMENDED.
The authorization server MUST implement TLS.
Which version(s) ought to be implemented will vary over
time, and depend on the widespread deployment and known
security vulnerabilities at the time of implementation.
At the time of this writing,
TLS version 1.2
is the most recent version, but has very limited actual
deployment, and might not be readily available in
implementation toolkits.
TLS version 1.0
is the most widely deployed version, and will give the
broadest interoperability.
To protect against token disclosure, confidentiality
protection MUST be applied using
TLS
with a ciphersuite that provides confidentiality and
integrity protection. This
requires that the communication interaction between the
client and the authorization server, as well as the
interaction between the client and the resource server,
utilize confidentiality and integrity protection.
Since TLS is mandatory to
implement and to use with this specification, it is the
preferred approach for preventing token disclosure via the
communication channel. For those cases where the client
is prevented from observing the contents of the token, token
encryption MUST be applied in addition to the usage of TLS
protection.
As a further defense against token disclosure, the client
MUST validate the TLS certificate chain when making requests
to protected resources.
Cookies are typically transmitted in the clear. Thus, any
information contained in them is at risk of disclosure.
Therefore, bearer tokens MUST NOT be stored in cookies that
can be sent in the clear.
In some deployments, including those utilizing load
balancers, the TLS connection to the resource server
terminates prior to the actual server that provides the
resource. This could leave the token unprotected between
the front end server where the TLS connection terminates and
the back end server that provides the resource. In such
deployments, sufficient measures MUST be employed to ensure
confidentiality of the token between the front end and
back end servers; encryption of the token is one possible
such measure.
To deal with token capture and replay,
the following recommendations are
made: First, the lifetime of the token MUST be limited;
one means of achieving this is by
putting a validity time field inside the protected part of
the token. Note that using short-lived (one hour or less)
tokens reduces the impact of them being
leaked. Second, confidentiality protection of the exchanges
between the client and the authorization server and between
the client and the resource server MUST be applied.
As a
consequence, no eavesdropper along the communication path is
able to observe the token exchange. Consequently, such an
on-path adversary cannot replay the token.
Furthermore, when
presenting the token to a resource server, the client MUST
verify the identity of that resource server, as per
Representation and Verification of Domain-Based Application Service
Identity within Internet Public Key Infrastructure Using X.509 (PKIX)
Certificates in the Context of Transport Layer Security (TLS)
.
Note that the
client MUST validate the TLS certificate chain when making
these requests to protected resources. Presenting the token
to an unauthenticated and unauthorized resource server or
failing to validate the certificate chain will allow
adversaries to steal the token and gain unauthorized access
to protected resources.
Client implementations MUST ensure that bearer tokens
are not leaked to unintended parties, as they will be
able to use them to gain access to protected resources.
This is the primary security consideration when using
bearer tokens and underlies all the more
specific recommendations that follow.
The client MUST validate the TLS certificate chain when
making requests to protected resources. Failing to do
so may enable DNS hijacking attacks to steal the token
and gain unintended access.
Clients MUST always use
TLS
(https) or equivalent transport security when making requests
with bearer tokens. Failing to do so exposes the token
to numerous attacks that could give attackers unintended
access.
Implementations MUST NOT store bearer tokens within
cookies that can be sent in the clear (which is the
default transmission mode for cookies).
Implementations that do store bearer tokens in cookies
MUST take precautions against cross site request forgery.
Token servers SHOULD issue short-lived (one hour or
less) bearer tokens, particularly when issuing tokens to
clients that run within a web browser or other
environments where information leakage may occur. Using
short-lived bearer tokens can reduce the impact of them
being leaked.
Token servers SHOULD issue bearer tokens that contain an audience
restriction, scoping their use to the intended relying
party or set of relying parties.
Bearer tokens SHOULD NOT be passed in page URLs (for
example as query string parameters). Instead, bearer
tokens SHOULD be passed in HTTP message headers or
message bodies for which confidentiality measures are
taken. Browsers, web servers, and other software may not
adequately secure URLs in the browser history, web
server logs, and other data structures. If bearer tokens
are passed in page URLs, attackers might be able to
steal them from the history data, logs, or other
unsecured locations.
This specification registers the following access token type in the OAuth Access Token
Type Registry.
Bearer
(none)
Bearer
IETF
[[ this document ]]
This specification registers the following authentication
scheme in the Authentication Scheme Registry defined in
HTTP/1.1, Part 7 .
Bearer
[[ this document ]]
(none)
Coded Character Set -- 7-bit American Standard Code for Information InterchangeAmerican National Standards InstituteNIST Special Publication 800-63-1, INFORMATION SECURITYNISTNISTNISTNISTNISTNIST
The following people contributed to preliminary versions of this document:
Blaine Cook (BT), Brian Eaton (Google), Yaron Y. Goland (Microsoft), Brent Goldman (Facebook),
Raffi Krikorian (Twitter), Luke Shepard (Facebook), and Allen Tom (Yahoo!). The content and
concepts within are a product of the OAuth community, the WRAP community, and the OAuth Working
Group.
The OAuth Working Group has dozens of very active contributors who proposed ideas and
wording for this document, including:
Michael Adams, Amanda Anganes, Andrew Arnott, Dirk Balfanz,
John Bradley, Brian Campbell, Leah Culver, Bill de hÓra,
Brian Ellin, Igor Faynberg, Stephen Farrell, George Fletcher,
Tim Freeman, Evan Gilbert, Yaron Y. Goland, Thomas Hardjono,
Justin Hart, Phil Hunt, John Kemp, Eran Hammer-Lahav,
Chasen Le Hara, Barry Leiba, Michael B. Jones,
Torsten Lodderstedt, Eve Maler, James Manger, Laurence Miao,
William J. Mills, Chuck Mortimore, Anthony Nadalin,
Julian Reschke, Justin Richer, Peter Saint-Andre, Nat Sakimura,
Rob Sayre, Marius Scurtescu, Naitik Shah, Justin Smith,
Jeremy Suriel, Christian Stübner, Paul Tarjan,
Hannes Tschofenig, Franklin Tse, and Shane Weeden.
[[ to be removed by the RFC editor before publication as an RFC ]]
-16
Use the HTTPbis auth-param syntax for Bearer challenge
attributes.
Dropped the sentence "The realm value is intended for
programmatic use and is not meant to be displayed to end
users".
Reordered form-encoded body parameter description bullets
for better readability.
Added reference.
-15
Clarified that form-encoded content must consist entirely
of ASCII characters.
Added TLS version requirements.
Applied editorial improvements suggested by Mark
Nottingham during the APPS area review.
-14
Changes made in response to review comments by Security
Area Director Stephen Farrell. Specifically:
Strengthened warnings about passing an access token as a
query parameter and more precisely described the
limitations placed upon the use of this method.
Clarified that the realm
attribute MAY included to indicate the scope of protection
in the manner described in
HTTP/1.1, Part 7 .
Normatively stated that "the token integrity protection
MUST be sufficient to prevent the token from being
modified".
Added statement that "TLS is mandatory to implement and
use with this specification" to the introduction.
Stated that TLS MUST be used with "a ciphersuite that
provides confidentiality and integrity protection".
Added "As a further defense against token disclosure, the
client MUST validate the TLS certificate chain when making
requests to protected resources" to the Threat Mitigation
section.
Clarified that putting a validity time field inside the
protected part of the token is one means, but not the only
means, of limiting the lifetime of the token.
Dropped the confusing phrase "for instance, through the
use of TLS" from the sentence about confidentiality
protection of the exchanges.
Reference RFC 6125 for identity verification, rather than
RFC 2818.
Stated that the token MUST be protected between front end
and back end servers when the TLS connection terminates at
a front end server that is distinct from the actual server
that provides the resource.
Stated that bearer tokens MUST NOT be stored in cookies
that can be sent in the clear in the Threat Mitigation
section.
Replaced sole remaining reference to with
HTTPbis
reference.
Replaced all references where the reference is used as if
it were part of the sentence (such as "defined by
[I-D.whatever]") with ones where the specification name is
used, followed by the reference (such as "defined by
Whatever [I-D.whatever]").
Other on-normative editorial improvements.
-13
At the request of Hannes Tschofenig, made ABNF changes to
make it clear that no special WWW-Authenticate response
header field parsers are needed. The scope, error-description, and error-uri parameters are all now
defined as quoted-string in the ABNF (as error already was). Restrictions on
these values that were formerly described in the ABNFs are
now described in normative text instead.
-12
Made non-normative editorial changes that Hannes
Tschofenig requested be applied prior to forwarding the
specification to the IESG.
Added rationale for the choice of the b64token syntax.
Added rationale stating that receivers are free to parse
the scope attribute using a
standard quoted-string parser, since it will correctly
process all legal scope
values.
Added additional active working group contributors to the
Acknowledgements section.
-11
Replaced uses of <"> with DQUOTE to pass ABNF syntax check.
-10
Removed the #auth-param option from Authorization header
syntax (leaving only the b64token syntax).
Restricted the scope value
character set to %x21 / %x23-5B / %x5D-7E (printable ASCII
characters excluding double-quote and backslash).
Indicated that scope is intended for programmatic use and
is not meant to be displayed to end users.
Restricted the character set for error_description strings to SP /
VCHAR and indicated that they are not meant to be
displayed to end users.
Included more description in the Abstract, since Hannes
Tschofenig indicated that the RFC editor would require
this.
Changed "Access Grant" to "Authorization Grant", as was
done in the core spec.
Simplified the introduction to the Authenticated Requests
section.
-09
Incorporated working group last call comments. Specific changes were:
Use definitions from rather than .
Update credentials definition to conform to .
Further clarified that query parameters may occur in any order.
Specify that error_description is UTF-8 encoded
(matching the core specification).
Registered "Bearer" Authentication Scheme in
Authentication Scheme Registry defined by
.
Updated references to oauth-v2, httpbis-p1-messaging, and
httpbis-p7-auth drafts.
Other wording improvements not introducing normative changes.
-08
Updated references to oauth-v2 and HTTPbis drafts.
-07
Added missing comma in error response example.
-06
Changed parameter name bearer_token to access_token, per working group
consensus.
Changed HTTP status code for invalid_request error code from HTTP
401 (Unauthorized) back to HTTP 400 (Bad Request), per
input from HTTP working group experts.
-05
Removed OAuth Errors Registry, per design team input.
Changed HTTP status code for invalid_request error code from HTTP
400 (Bad Request) to HTTP 401 (Unauthorized) to match HTTP
usage [[ change pending working group consensus ]].
Added missing quotation marks in error-uri definition.
Added note to add language and encoding information to
error_description if the core specification does.
Explicitly reference the Augmented Backus-Naur Form (ABNF)
defined in .
Use auth-param instead of repeating its definition, which
is ( token "=" ( token / quoted-string ) ).
Clarify security considerations about including an
audience restriction in the token and include a
recommendation to issue scoped bearer tokens in the
summary of recommendations.
-04
Edits responding to working group last call feedback on
-03. Specific edits enumerated below.
Added Bearer Token definition in Terminology section.
Changed parameter name oauth_token to bearer_token.
Added realm parameter to WWW-Authenticate response to comply
with .
Removed "[ RWS 1#auth-param ]" from credentials definition since it did
not comply with the ABNF in .
Removed restriction that the bearer_token (formerly oauth_token) parameter be the last
parameter in the entity-body and the HTTP request URI
query.
Do not require WWW-Authenticate Response in a reply to a
malformed request, as an HTTP 400 Bad Request response
without a WWW-Authenticate header is likely the right
response in some cases of malformed requests.
Removed OAuth Parameters registry extension.
Numerous editorial improvements suggested by working group
members.
-03
Restored the WWW-Authenticate response header
functionality deleted from the framework specification in
draft 12 based upon the specification text from draft 11.
Augmented the OAuth Parameters registry by adding two
additional parameter usage locations: "resource request"
and "resource response".
Registered the "oauth_token" OAuth parameter with usage
location "resource request".
Registered the "error" OAuth parameter.
Created the OAuth Error registry and registered errors.
Changed the "OAuth2" OAuth access token type name to
"Bearer".
-02
Incorporated feedback received on draft 01. Most changes
were to the security considerations section. No normative
changes were made. Specific changes included:
Changed terminology from "token reuse" to "token capture
and replay".
Removed sentence "Encrypting the token contents is another
alternative" from the security considerations since it was
redundant and potentially confusing.
Corrected some references to "resource server" to be
"authorization server" in the security considerations.
Generalized security considerations language about
obtaining consent of the resource owner.
Broadened scope of security considerations description for
recommendation "Don't pass bearer tokens in page URLs".
Removed unused reference to OAuth 1.0.
Updated reference to framework specification and updated
David Recordon's e-mail address.
Removed security considerations text on authenticating
clients.
Registered the "OAuth2" OAuth access token type and
"oauth_token" parameter.
-01
First public draft, which incorporates feedback received
on -00 including enhanced Security Considerations content.
This version is intended to accompany OAuth 2.0 draft 11.
-00
Initial draft based on preliminary version of OAuth 2.0 draft 11.