oauth2 2.0.18 → 2.0.22
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +4 -4
- checksums.yaml.gz.sig +0 -0
- data/CHANGELOG.md +138 -5
- data/CITATION.cff +6 -6
- data/CONTRIBUTING.md +82 -31
- data/FUNDING.md +1 -1
- data/LICENSE.md +110 -0
- data/README.md +341 -734
- data/SECURITY.md +1 -4
- data/certs/pboling.pem +27 -0
- data/lib/oauth2/access_token.rb +11 -14
- data/lib/oauth2/auth_sanitizer.rb +36 -0
- data/lib/oauth2/authenticator.rb +9 -7
- data/lib/oauth2/client.rb +46 -5
- data/lib/oauth2/error.rb +2 -0
- data/lib/oauth2/filtered_attributes.rb +7 -49
- data/lib/oauth2/response.rb +14 -12
- data/lib/oauth2/version.rb +2 -1
- data/lib/oauth2.rb +39 -17
- data/sig/oauth2/filtered_attributes.rbs +6 -1
- data/sig/oauth2/sanitized_logger.rbs +32 -0
- data/sig/oauth2/thing_filter.rbs +10 -0
- data/sig/oauth2/version.rbs +1 -0
- data.tar.gz.sig +0 -0
- metadata +124 -103
- metadata.gz.sig +0 -0
- data/IRP.md +0 -107
- data/LICENSE.txt +0 -22
- data/OIDC.md +0 -167
- data/REEK +0 -0
- data/THREAT_MODEL.md +0 -85
data/OIDC.md
DELETED
|
@@ -1,167 +0,0 @@
|
|
|
1
|
-
# OpenID Connect (OIDC) with ruby-oauth/oauth2
|
|
2
|
-
|
|
3
|
-
## OIDC Libraries
|
|
4
|
-
|
|
5
|
-
Libraries built on top of the oauth2 gem that implement OIDC.
|
|
6
|
-
|
|
7
|
-
- [gamora](https://github.com/amco/gamora-rb) - OpenID Connect Relying Party for Rails apps
|
|
8
|
-
- [omniauth-doximity-oauth2](https://github.com/doximity/omniauth-doximity-oauth2) - OmniAuth strategy for Doximity, supporting OIDC, and using PKCE
|
|
9
|
-
- [omniauth-himari](https://github.com/sorah/himari) - OmniAuth strategy to act as OIDC RP and use [Himari](https://github.com/sorah/himari) for OP
|
|
10
|
-
- [omniauth-mit-oauth2](https://github.com/MITLibraries/omniauth-mit-oauth2) - OmniAuth strategy for MIT OIDC
|
|
11
|
-
|
|
12
|
-
If any other libraries would like to be added to this list, please open an issue or pull request.
|
|
13
|
-
|
|
14
|
-
## Raw OIDC with ruby-oauth/oauth2
|
|
15
|
-
|
|
16
|
-
This document complements the inline documentation by focusing on OpenID Connect (OIDC) 1.0 usage patterns when using this gem as an OAuth 2.0 client library.
|
|
17
|
-
|
|
18
|
-
Scope of this document
|
|
19
|
-
|
|
20
|
-
- Audience: Developers building an OAuth 2.0/OIDC Relying Party (RP, aka client) in Ruby.
|
|
21
|
-
- Non-goals: This gem does not implement an OIDC Provider (OP, aka Authorization Server); for OP/server see other projects (e.g., doorkeeper + oidc extensions).
|
|
22
|
-
- Status: Informational documentation with links to normative specs. The gem intentionally remains protocol-agnostic beyond OAuth 2.0; OIDC specifics (like ID Token validation) must be handled by your application.
|
|
23
|
-
|
|
24
|
-
Key concepts refresher
|
|
25
|
-
|
|
26
|
-
- OAuth 2.0 delegates authorization; it does not define authentication of the end-user.
|
|
27
|
-
- OIDC layers an identity layer on top of OAuth 2.0, introducing:
|
|
28
|
-
- ID Token: a JWT carrying claims about the authenticated end-user and the authentication event.
|
|
29
|
-
- Standardized scopes: openid (mandatory), profile, email, address, phone, offline_access, and others.
|
|
30
|
-
- UserInfo endpoint: a protected resource for retrieving user profile claims.
|
|
31
|
-
- Discovery and Dynamic Client Registration (optional for providers/clients that support them).
|
|
32
|
-
|
|
33
|
-
What this gem provides for OIDC
|
|
34
|
-
|
|
35
|
-
- All OAuth 2.0 client capabilities required for OIDC flows: building authorization requests, exchanging authorization codes, refreshing tokens, and making authenticated resource requests.
|
|
36
|
-
- Transport and parsing conveniences (snaky hash, Faraday integration, error handling, etc.).
|
|
37
|
-
- Optional client authentication schemes useful with OIDC deployments:
|
|
38
|
-
- basic_auth (default)
|
|
39
|
-
- request_body (legacy)
|
|
40
|
-
- tls_client_auth (MTLS)
|
|
41
|
-
- private_key_jwt (OIDC-compliant when configured per OP requirements)
|
|
42
|
-
|
|
43
|
-
What you must add in your app for OIDC
|
|
44
|
-
|
|
45
|
-
- ID Token validation: This gem surfaces id_token values but does not verify them. Your app should:
|
|
46
|
-
1) Parse the JWT (header, payload, signature)
|
|
47
|
-
2) Fetch the OP JSON Web Key Set (JWKS) from discovery (or configure statically)
|
|
48
|
-
3) Select the correct key by kid (when present) and verify the signature and algorithm
|
|
49
|
-
4) Validate standard claims (iss, aud, exp, iat, nbf, azp, nonce when used, at_hash/c_hash when applicable)
|
|
50
|
-
5) Enforce expected client_id, issuer, and clock skew policies
|
|
51
|
-
- Nonce handling for Authorization Code flow with OIDC: generate a cryptographically-random nonce, bind it to the user session before redirect, include it in authorize request, and verify it in the ID Token on return.
|
|
52
|
-
- PKCE is best practice and often required by OPs: generate/verifier, send challenge in authorize, send verifier in token request.
|
|
53
|
-
- Session/state management: continue to validate state to mitigate CSRF; use exact redirect_uri matching.
|
|
54
|
-
|
|
55
|
-
Minimal OIDC Authorization Code example
|
|
56
|
-
|
|
57
|
-
```ruby
|
|
58
|
-
require "oauth2"
|
|
59
|
-
require "jwt" # jwt/ruby-jwt
|
|
60
|
-
require "net/http"
|
|
61
|
-
require "json"
|
|
62
|
-
|
|
63
|
-
client = OAuth2::Client.new(
|
|
64
|
-
ENV.fetch("OIDC_CLIENT_ID"),
|
|
65
|
-
ENV.fetch("OIDC_CLIENT_SECRET"),
|
|
66
|
-
site: ENV.fetch("OIDC_ISSUER"), # e.g. https://accounts.example.com
|
|
67
|
-
authorize_url: "/authorize", # or discovered
|
|
68
|
-
token_url: "/token", # or discovered
|
|
69
|
-
)
|
|
70
|
-
|
|
71
|
-
# Step 1: Redirect to OP for consent/auth
|
|
72
|
-
state = SecureRandom.hex(16)
|
|
73
|
-
nonce = SecureRandom.hex(16)
|
|
74
|
-
pkce_verifier = SecureRandom.urlsafe_base64(64)
|
|
75
|
-
pkce_challenge = Base64.urlsafe_encode64(Digest::SHA256.digest(pkce_verifier)).delete("=")
|
|
76
|
-
|
|
77
|
-
authz_url = client.auth_code.authorize_url(
|
|
78
|
-
scope: "openid profile email",
|
|
79
|
-
state: state,
|
|
80
|
-
nonce: nonce,
|
|
81
|
-
code_challenge: pkce_challenge,
|
|
82
|
-
code_challenge_method: "S256",
|
|
83
|
-
redirect_uri: ENV.fetch("OIDC_REDIRECT_URI"),
|
|
84
|
-
)
|
|
85
|
-
# redirect_to authz_url
|
|
86
|
-
|
|
87
|
-
# Step 2: Handle callback
|
|
88
|
-
# params[:code], params[:state]
|
|
89
|
-
raise "state mismatch" unless params[:state] == state
|
|
90
|
-
|
|
91
|
-
token = client.auth_code.get_token(
|
|
92
|
-
params[:code],
|
|
93
|
-
redirect_uri: ENV.fetch("OIDC_REDIRECT_URI"),
|
|
94
|
-
code_verifier: pkce_verifier,
|
|
95
|
-
)
|
|
96
|
-
|
|
97
|
-
# The token may include: access_token, id_token, refresh_token, etc.
|
|
98
|
-
id_token = token.params["id_token"] || token.params[:id_token]
|
|
99
|
-
|
|
100
|
-
# Step 3: Validate the ID Token (simplified – add your own checks!)
|
|
101
|
-
# Discover keys (example using .well-known)
|
|
102
|
-
issuer = ENV.fetch("OIDC_ISSUER")
|
|
103
|
-
jwks_uri = JSON.parse(Net::HTTP.get(URI.join(issuer, "/.well-known/openid-configuration"))).
|
|
104
|
-
fetch("jwks_uri")
|
|
105
|
-
jwks = JSON.parse(Net::HTTP.get(URI(jwks_uri)))
|
|
106
|
-
keys = jwks.fetch("keys")
|
|
107
|
-
|
|
108
|
-
# Use ruby-jwt JWK loader
|
|
109
|
-
jwk_set = JWT::JWK::Set.new(keys.map { |k| JWT::JWK.import(k) })
|
|
110
|
-
|
|
111
|
-
decoded, headers = JWT.decode(
|
|
112
|
-
id_token,
|
|
113
|
-
nil,
|
|
114
|
-
true,
|
|
115
|
-
algorithms: ["RS256", "ES256", "PS256"],
|
|
116
|
-
jwks: jwk_set,
|
|
117
|
-
verify_iss: true,
|
|
118
|
-
iss: issuer,
|
|
119
|
-
verify_aud: true,
|
|
120
|
-
aud: ENV.fetch("OIDC_CLIENT_ID"),
|
|
121
|
-
)
|
|
122
|
-
|
|
123
|
-
# Verify nonce
|
|
124
|
-
raise "nonce mismatch" unless decoded["nonce"] == nonce
|
|
125
|
-
|
|
126
|
-
# Optionally: call UserInfo
|
|
127
|
-
userinfo = token.get("/userinfo").parsed
|
|
128
|
-
```
|
|
129
|
-
|
|
130
|
-
Notes on discovery and registration
|
|
131
|
-
|
|
132
|
-
- Discovery: Most OPs publish configuration at `{issuer}/.well-known/openid-configuration` (OIDC Discovery 1.0). From there, resolve authorization_endpoint, token_endpoint, jwks_uri, userinfo_endpoint, etc.
|
|
133
|
-
- Dynamic Client Registration: Some OPs allow registering clients programmatically (OIDC Dynamic Client Registration 1.0). This gem does not implement registration; use a plain HTTP client or Faraday and store credentials securely.
|
|
134
|
-
|
|
135
|
-
Common pitfalls and tips
|
|
136
|
-
|
|
137
|
-
- Always request the openid scope when you expect an ID Token. Without it, the OP may behave as vanilla OAuth 2.0.
|
|
138
|
-
- Validate ID Token signature and claims before trusting any identity data. Do not rely solely on the presence of an id_token field.
|
|
139
|
-
- Prefer Authorization Code + PKCE. Avoid Implicit; it is discouraged in modern guidance and may be disabled by providers.
|
|
140
|
-
- Use exact redirect_uri matching, and keep your allow-list short.
|
|
141
|
-
- For public clients that use refresh tokens, prefer sender-constrained tokens (DPoP/MTLS) or rotation with one-time-use refresh tokens, per modern best practices.
|
|
142
|
-
- When using private_key_jwt, ensure the "aud" (or token_url) and "iss/sub" claims are set per the OP’s rules, and include kid in the JWT header when required so the OP can select the right key.
|
|
143
|
-
|
|
144
|
-
Relevant specifications and references
|
|
145
|
-
|
|
146
|
-
- OpenID Connect Core 1.0: https://openid.net/specs/openid-connect-core-1_0.html
|
|
147
|
-
- OIDC Core (final): https://openid.net/specs/openid-connect-core-1_0-final.html
|
|
148
|
-
- How OIDC works: https://openid.net/developers/how-connect-works/
|
|
149
|
-
- OpenID Connect home: https://openid.net/connect/
|
|
150
|
-
- OIDC Discovery 1.0: https://openid.net/specs/openid-connect-discovery-1_0.html
|
|
151
|
-
- OIDC Dynamic Client Registration 1.0: https://openid.net/specs/openid-connect-registration-1_0.html
|
|
152
|
-
- OIDC Session Management 1.0: https://openid.net/specs/openid-connect-session-1_0.html
|
|
153
|
-
- OIDC RP-Initiated Logout 1.0: https://openid.net/specs/openid-connect-rpinitiated-1_0.html
|
|
154
|
-
- OIDC Back-Channel Logout 1.0: https://openid.net/specs/openid-connect-backchannel-1_0.html
|
|
155
|
-
- OIDC Front-Channel Logout 1.0: https://openid.net/specs/openid-connect-frontchannel-1_0.html
|
|
156
|
-
- Auth0 OIDC overview: https://auth0.com/docs/authenticate/protocols/openid-connect-protocol
|
|
157
|
-
- Spring Authorization Server’s list of OAuth2/OIDC specs: https://github.com/spring-projects/spring-authorization-server/wiki/OAuth2-and-OIDC-Specifications
|
|
158
|
-
|
|
159
|
-
See also
|
|
160
|
-
|
|
161
|
-
- README sections on OAuth 2.1 notes and OIDC notes
|
|
162
|
-
- Strategy classes under lib/oauth2/strategy for flow helpers
|
|
163
|
-
- Specs under spec/oauth2 for concrete usage patterns
|
|
164
|
-
|
|
165
|
-
Contributions welcome
|
|
166
|
-
|
|
167
|
-
- If you discover provider-specific nuances, consider contributing examples or clarifications (without embedding provider-specific hacks into the library).
|
data/REEK
DELETED
|
File without changes
|
data/THREAT_MODEL.md
DELETED
|
@@ -1,85 +0,0 @@
|
|
|
1
|
-
# Threat Model Outline for oauth2 Ruby Gem
|
|
2
|
-
|
|
3
|
-
## 1. Overview
|
|
4
|
-
This document outlines the threat model for the `oauth2` Ruby gem, which implements OAuth 2.0, 2.1, and OIDC Core protocols. The gem is used to facilitate secure authorization and authentication in Ruby applications.
|
|
5
|
-
|
|
6
|
-
## 2. Assets to Protect
|
|
7
|
-
- OAuth access tokens, refresh tokens, and ID tokens
|
|
8
|
-
- User credentials (if handled)
|
|
9
|
-
- Client secrets and application credentials
|
|
10
|
-
- Sensitive user data accessed via OAuth
|
|
11
|
-
- Private keys and certificates (for signing/verifying tokens)
|
|
12
|
-
|
|
13
|
-
## 3. Potential Threat Actors
|
|
14
|
-
- External attackers (internet-based)
|
|
15
|
-
- Malicious OAuth clients or resource servers
|
|
16
|
-
- Insiders (developers, maintainers)
|
|
17
|
-
- Compromised dependencies
|
|
18
|
-
|
|
19
|
-
## 4. Attack Surfaces
|
|
20
|
-
- OAuth endpoints (authorization, token, revocation, introspection)
|
|
21
|
-
- HTTP request/response handling
|
|
22
|
-
- Token storage and management
|
|
23
|
-
- Configuration files and environment variables
|
|
24
|
-
- Dependency supply chain
|
|
25
|
-
|
|
26
|
-
## 5. Threats and Mitigations
|
|
27
|
-
|
|
28
|
-
### 5.1 Token Leakage
|
|
29
|
-
- **Threat:** Tokens exposed via logs, URLs, or insecure storage
|
|
30
|
-
- **Mitigations:**
|
|
31
|
-
- Avoid logging sensitive tokens
|
|
32
|
-
- Use secure storage mechanisms
|
|
33
|
-
- Never expose tokens in URLs
|
|
34
|
-
|
|
35
|
-
### 5.2 Token Replay and Forgery
|
|
36
|
-
- **Threat:** Attackers reuse or forge tokens
|
|
37
|
-
- **Mitigations:**
|
|
38
|
-
- Validate token signatures and claims
|
|
39
|
-
- Use short-lived tokens and refresh tokens
|
|
40
|
-
- Implement token revocation
|
|
41
|
-
|
|
42
|
-
### 5.3 Insecure Communication
|
|
43
|
-
- **Threat:** Data intercepted via MITM attacks
|
|
44
|
-
- **Mitigations:**
|
|
45
|
-
- Enforce HTTPS for all communications
|
|
46
|
-
- Validate SSL/TLS certificates
|
|
47
|
-
|
|
48
|
-
### 5.4 Client Secret Exposure
|
|
49
|
-
- **Threat:** Client secrets leaked in code or version control
|
|
50
|
-
- **Mitigations:**
|
|
51
|
-
- Store secrets in environment variables or secure vaults
|
|
52
|
-
- Never commit secrets to source control
|
|
53
|
-
|
|
54
|
-
### 5.5 Dependency Vulnerabilities
|
|
55
|
-
- **Threat:** Vulnerabilities in third-party libraries
|
|
56
|
-
- **Mitigations:**
|
|
57
|
-
- Regularly update dependencies
|
|
58
|
-
- Use tools like `bundler-audit` for vulnerability scanning
|
|
59
|
-
|
|
60
|
-
### 5.6 Improper Input Validation
|
|
61
|
-
- **Threat:** Injection attacks via untrusted input
|
|
62
|
-
- **Mitigations:**
|
|
63
|
-
- Validate and sanitize all inputs
|
|
64
|
-
- Use parameterized queries and safe APIs
|
|
65
|
-
|
|
66
|
-
### 5.7 Insufficient Logging and Monitoring
|
|
67
|
-
- **Threat:** Attacks go undetected
|
|
68
|
-
- **Mitigations:**
|
|
69
|
-
- Log security-relevant events (without sensitive data)
|
|
70
|
-
- Monitor for suspicious activity
|
|
71
|
-
|
|
72
|
-
## 6. Assumptions
|
|
73
|
-
- The gem is used in a secure environment with up-to-date Ruby and dependencies
|
|
74
|
-
- End-users are responsible for secure configuration and deployment
|
|
75
|
-
|
|
76
|
-
## 7. Out of Scope
|
|
77
|
-
- Security of external OAuth providers
|
|
78
|
-
- Application-level business logic
|
|
79
|
-
|
|
80
|
-
## 8. References
|
|
81
|
-
- [OAuth 2.0 Threat Model and Security Considerations (RFC 6819)](https://tools.ietf.org/html/rfc6819)
|
|
82
|
-
- [OWASP Top Ten](https://owasp.org/www-project-top-ten/)
|
|
83
|
-
|
|
84
|
-
---
|
|
85
|
-
This outline should be reviewed and updated regularly as the project evolves.
|