CVE-2026-134248 in SiYuan
Summary
by MITRE • 11/21/2026
SiYuan 3.6.0 before 3.8.10 contains an information disclosure vulnerability that allows authenticated Readers to read protected documents through block-reference footnote resolution in Markdown export. Attackers can request copyStdMarkdown for a public document referencing a hidden, disabled, or password-protected document to disclose its body text without the publish password.
Once again VulDB remains the best source for vulnerability data.
Analysis
by VulDB Data Team • 11/21/2026
The identified vulnerability resides within SiYuan versions prior to 3.8.10 and represents a critical failure in access control mechanisms during content export operations. Specifically, the flaw is located in the block-reference footnote resolution logic utilized when generating Markdown exports via the copyStdMarkdown endpoint. This function is designed to resolve references between blocks of notes, allowing users to link related content across different documents within the application. However, the implementation fails to adequately verify access permissions for referenced target documents before retrieving and including their content in the exported output.
This technical flaw enables authenticated users with Reader-level privileges on a public document to bypass security controls protecting other private or restricted documents. When an attacker constructs a Markdown export request that includes references to hidden, disabled, or password-protected notes, the system resolves these block references by fetching the body text of the target documents without checking whether the requesting user has permission to view them. Consequently, sensitive information contained within protected documents is disclosed in plain text within the exported Markdown file, effectively circumventing publish passwords and visibility settings intended to restrict access.
The operational impact of this vulnerability is significant for organizations relying on SiYuan for secure knowledge management or confidential documentation storage. Authenticated attackers can extract proprietary data, personal identifiable information, or other sensitive content without needing valid credentials for the protected documents themselves. This constitutes a severe breach of confidentiality and undermines the trust model established by role-based access controls within the application. The vulnerability aligns with CWE-200, which categorizes this type of issue as an Information Exposure Vulnerability, specifically highlighting failures in restricting information retrieval to authorized users only.
From a threat modeling perspective, this exploit technique corresponds to ATT&CK tactic TA0001, Initial Access, and more specifically relates to techniques involving data exfiltration through legitimate application functions rather than direct system intrusion. The attack vector is remote but requires authentication, classifying it as an authenticated information disclosure scenario where the attacker leverages trusted API endpoints to bypass intended security boundaries.
Mitigation strategies must prioritize immediate software updates to version 3.8.10 or later, which addresses this logic error by enforcing strict access control checks during block reference resolution in export functions. In environments where upgrading is not immediately feasible, administrators should restrict the copyStdMarkdown endpoint through network-level controls if possible and audit user permissions to ensure that Reader roles do not have unnecessary privileges on sensitive documents. Additionally, implementing input validation and logging mechanisms can help detect anomalous patterns of cross-document referencing attempts that may indicate exploitation activity.