/

CVE-2026-72793

Severity:

HIGH RISK

(8.6)

Sensitive Data Exposure (CWE-522) in SiYuan System Endpoint

CVE-2026-72793

Deferred

Vulnerability type

Category:

Sensitive Data Exposure

CWE:

CWE-522

Impact:

Credential Disclosure and Privilege Escalation

Severity:

HIGH RISK

(8.6)

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

SiYuan versions before 3.7.4 expose sensitive configuration values through the /api/system/getConf endpoint to anonymous or publish-reader users. The exposed data may include the session-cookie signing key, operating-system username, and encrypted-notebook key material, enabling cookie forgery and possible administrator impersonation when access-auth codes are not configured.

Affected context

The issue occurs in SiYuan deployments where the system configuration endpoint is reachable by unauthenticated users or users with publish-reader access and sensitive configuration fields are returned without adequate masking. Risk is higher when the deployment does not use access-auth codes, because a disclosed session-signing key may support administrator-level impersonation.

Exploit preconditions

  • Attack surface: The /api/system/getConf HTTP API endpoint
  • Required access: Anonymous access or publish-reader permissions, depending on deployment configuration
  • Trigger point: Requesting system configuration data from the endpoint
  • Why exploitation is possible: Sensitive secrets and key material are returned to callers instead of being consistently redacted, allowing exposed authentication-signing material to be reused

Why this happens

The underlying bug is insufficient protection of sensitive information in an API response. Configuration serialization exposes secrets that should be omitted or masked according to the caller's privilege and the field's sensitivity. This maps to CWE-522, where credentials or authentication material are inadequately protected and can be disclosed to an unauthorized party.

Potential impact

  • Attackers may forge or tamper with session cookies
  • Unauthorized users could impersonate valid accounts
  • Deployments without access-auth codes may allow administrator impersonation
  • Encrypted-notebook key material could be exposed
  • The operating-system username and related deployment details may be disclosed
  • Attackers could gain access to protected content or alter application data

General mitigation guidance

Mitigation generally focuses on reducing the amount of sensitive data returned

by configuration APIs and enforcing least privilege.

Common defensive patterns include:

  • Excluding session-signing keys, encryption keys, and other secrets from API responses by default
  • Applying field-level redaction based on sensitivity and caller privilege
  • Requiring strong authentication and authorization for diagnostic or configuration endpoints
  • Treating publish-reader access as untrusted for secret disclosure decisions
  • Configuring an additional access-auth control where appropriate
  • Rotating potentially exposed signing keys and encryption material through established secret-management processes
  • Running the application with the minimum required operating-system permissions

How teams detect issues like this

Organizations can assess this issue through:

  • API tests that request configuration endpoints using anonymous and low-privilege identities
  • Response-schema checks that fail when secret-like fields appear in externally reachable responses
  • Static analysis and code review for configuration serializers that return complete internal objects
  • Monitoring for unusual access to system-configuration endpoints and anomalous session activity
  • Policy enforcement requiring explicit allowlists for fields exposed by administrative APIs
  • Automated security review tools that identify credential, key, or token material in API responses

Sources

SourceLink
github.comgithub.com/siyuan-note/siyuan/security/advisories/GHSA-h4v5-crx2-3cv4
vulncheck.comvulncheck.com/advisories/siyuan-before-information-disclosure-via-api-system-getconf

Frequently asked

questions

Exploitability depends on deployment settings, but the endpoint may disclose

sensitive values to anonymous users or accounts with publish-reader access.

A disclosed session-cookie signing key may allow attackers to forge or alter

cookies, potentially impersonating users. Administrator impersonation is more

plausible when access-auth codes are not configured.

Return only the minimum necessary configuration fields, omit secrets entirely,

and enforce authentication and field-level authorization for sensitive

endpoints.

Use explicit response allowlists, classify configuration fields by

sensitivity, test endpoints under each privilege level, and monitor for

secret material appearing in API responses.

Find out what's already
exploitable in your codebase.