/

CVE-2026-57485

Severity:

HIGH RISK

(8.5)

API Key Exposure (CWE-200) in Stirling-PDF Pipeline

CVE-2026-57485

Deferred

Vulnerability type

Category:

Sensitive Information Exposure

CWE:

CWE-200, CWE-522

Impact:

Internal API Impersonation and Unauthorized Information Access

Severity:

HIGH RISK

(8.5)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

Stirling-PDF exposes an internal backend API key to authenticated users with the ROLE_USER permission through its user API-key functionality. A user who obtains the key may impersonate the internal service account, bypass ordinary rate limits, and access internal information endpoints.

Affected context

The issue occurs in Stirling-PDF deployments using the pipeline processing endpoint before the relevant security correction. Pipeline subrequests receive the STIRLING-PDF-BACKEND-API-USER credential, while an authenticated regular user can retrieve that credential through the API-key endpoint. The risk is greater when internal endpoints rely on the service-account key for authorization rather than independently enforcing the caller's privileges.

Exploit preconditions

  • Attack surface: Authenticated Stirling-PDF web and API endpoints
  • Required access: A valid account with the ROLE_USER permission
  • Trigger point: Access to the user API-key functionality followed by requests using the exposed backend credential
  • Why exploitation is possible: A credential intended for internal service-to-service requests is disclosed to a lower-privileged user and can be reused to authenticate against internal endpoints

Why this happens

The underlying issue is an authorization and secret-disclosure design flaw. A backend API key used by pipeline subrequests is exposed through a user-accessible API path, allowing a normal authenticated user to obtain a credential with broader trust than the user's own session.

This maps primarily to CWE-200: Exposure of Sensitive Information to an Unauthorized Actor. It is also related to CWE-522: Insufficiently Protected Credentials, because a service credential is made available beyond its intended trust boundary.

Potential impact

  • An authenticated user may impersonate the internal backend service account
  • Attackers could bypass endpoint-specific rate limits
  • Unauthorized users may access internal request or load information
  • Sensitive operational metadata could be disclosed
  • The exposed credential could support broader abuse if additional trusted endpoints accept it

General mitigation guidance

Mitigation generally focuses on separating internal service credentials from user-accessible API functionality and enforcing authorization at every endpoint.

Common defensive patterns include:

  • Never returning internal service-account keys through user-facing API responses
  • Using short-lived, narrowly scoped credentials for service-to-service communication
  • Enforcing authorization based on the originating user's privileges rather than trusting a shared backend key alone
  • Applying independent rate limits and access checks to internal and administrative endpoints
  • Storing secrets outside application responses and protecting them with least-privilege access controls
  • Rotating credentials after possible disclosure and invalidating unnecessary shared keys

How teams detect issues like this

Organizations can look for this class of issue through:

  • Static analysis and code review checks for secrets passed into user-facing responses
  • Tracing whether internal API credentials are injected into requests derived from user input or sessions
  • Testing authorization boundaries between regular users, service accounts, and administrative endpoints
  • Monitoring for the same API key being used across unrelated users or from unexpected request paths
  • Policy checks that prohibit hard-coded or broadly scoped credentials in browser-accessible API contracts
  • Automated API security testing for privilege escalation, credential exposure, and rate-limit bypass behavior

Sources

SourceLink
github.comgithub.com/Stirling-Tools/Stirling-PDF/commit/de9625942bbc329fdefcfde161476a633a3b3213
github.comgithub.com/Stirling-Tools/Stirling-PDF/pull/6047
github.comgithub.com/Stirling-Tools/Stirling-PDF/releases/tag/v2.9.0
github.comgithub.com/Stirling-Tools/Stirling-PDF/security/advisories/GHSA-3xxh-mm3g-c9w5

Frequently asked

questions

Yes, where the affected Stirling-PDF API is reachable and an attacker has authenticated access. Exploitation does not require direct host access if the relevant endpoints are exposed through the application.

No. The described attack requires a valid account with the ROLE_USER permission, after which the exposed internal credential may provide broader access.

Keep internal service credentials out of user-accessible endpoints, use narrowly scoped credentials, and enforce authorization independently at every internal or sensitive endpoint.

Treat service-account keys as secrets, review trust boundaries during API design, test role separation and credential exposure in CI, and monitor for shared credentials being reused across user contexts.

Find out what's already
exploitable in your codebase.