Description

SiYuan versions before v3.8.2 fail to properly filter private attribute-view cell values in the getAttributeViewKeys endpoint. Publish readers can retrieve hidden KeyValues payloads from rows bound to inaccessible documents, exposing private database contents without authorization.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "6.5",
  "pubDate": "2026-09-05T13:18:14.443Z",
  "pubdate": "2026-09-05T13:18:14.443Z",
  "executiveSummary": "An improper authorization vulnerability exists in SiYuan versions prior to v3.8.2, specifically within the getAttributeViewKeys endpoint. The flaw allows unauthorized users, including publish readers, to access and retrieve private attribute-view cell values.\nThis vulnerability is classified as an Information Disclosure issue where the application fails to enforce access control checks when querying database metadata and contents. By manipulating requests to the vulnerable endpoint, an attacker can bypass intended security boundaries to extract sensitive KeyValues payloads associated with documents that are restricted or inaccessible to the reader.\nThe risk implication is significant as it permits the exposure of private database contents, potentially leading to the leakage of sensitive user data, private notes, or internal metadata. Exploitation does not require elevated privileges, as it leverages the existing functionality exposed to publish readers. The vulnerability stems from insufficient server-side input validation and inadequate authorization verification on the backend API.\nUsers and administrators of the SiYuan platform are exposed to unauthorized data extraction unless the software is updated to a patched version that correctly validates document ownership and visibility prior to returning attribute-view data.",
  "technicalDetails": "The vulnerability resides in the getAttributeViewKeys endpoint within SiYuan, which is responsible for fetching attribute-view metadata and associated cell values. The root cause of this security defect is an Authorization Bypass, where the endpoint fails to properly validate the requesting user's access rights against the document object model (DOM) or the specific database records requested.\nIn the affected versions (prior to v3.8.2), the application logic performs a look-up for attribute-view keys but neglects to filter out records that are bound to private or protected documents. Consequently, when a publish reader initiates a request to the endpoint, the server processes the query for all bound rows without verifying if the document associated with those rows is marked as 'private' or 'inaccessible' to the authenticated session or public guest context.\nThe attack flow proceeds as follows: First, an attacker identifies the getAttributeViewKeys endpoint as a target. Since this endpoint lacks granular access control, the attacker crafts a request designed to retrieve attribute-view data. The server, acting as a trusted backend service, iterates through the database rows tied to the requested view. Because the input filtering mechanism is deficient, the server includes the contents of rows from restricted documents in the response payload. The attacker then parses the returned JSON or serialized data structures to extract hidden KeyValues payloads, effectively gaining read-access to private database contents that should be sequestered behind authorization barriers.\nThis exposure is critical because it bypasses the standard SiYuan access control model. Even if a document is hidden or restricted via the application's UI, the API continues to leak the underlying attribute data. The payload behavior involves the extraction of structured data, which may contain sensitive keys and values that provide insight into the structure and content of private user notes. Post-exploitation, the attacker possesses the capability to enumerate and exfiltrate private database contents systematically. This vulnerability exists regardless of the network environment, as the flaw is internal to the application's API logic. The exploitation is trivial, requiring only standard HTTP request capabilities to target the vulnerable function, and it successfully circumvents the intended security policies governing document visibility in the SiYuan ecosystem."
}