Description

SiYuan before v3.7.4 contains an information disclosure vulnerability in the /api/tag/getTag endpoint that returns tag labels and occurrence counts from password-protected documents to unauthenticated readers. Attackers can enumerate tag vocabulary and internal terminology from password-protected documents by calling the tag endpoint without providing the document's publish password.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "5.8",
  "pubDate": "2026-08-12T20:17:50.690Z",
  "pubdate": "2026-08-12T20:17:50.690Z",
  "executiveSummary": "SiYuan before v3.7.4 suffers from an information disclosure vulnerability located within the /api/tag/getTag endpoint. This flaw allows unauthenticated remote attackers to retrieve sensitive metadata, specifically including tag labels and occurrence counts originating from password-protected documents. The primary risk implication is the unauthorized leakage of internal terminology, thematic structure, and proprietary vocabulary intended to be secured behind document-level authentication mechanisms. Attackers require no prior authentication, special privileges, or knowledge of the publish password to successfully query the vulnerable endpoint. Exploitation is performed by directly issuing HTTP requests to the exposed API route, circumventing the intended confidentiality controls enforced on encrypted or password-restricted content within the application architecture.",
  "technicalDetails": "The vulnerability resides in the backend routing and authorization logic handling the /api/tag/getTag endpoint in SiYuan before v3.7.4. The root cause is an insufficient access control validation failure, where the API aggregates and returns tag metadata across the entire workspace or document repository without verifying whether the requesting user possesses the necessary decryption keys or publish passwords for password-protected documents. When an unauthenticated client sends a request to the vulnerable /api/tag/getTag endpoint over the network, the underlying function executes a database query or internal index lookup to retrieve all associated tag labels and their corresponding occurrence counts. Because the application fails to filter out tags derived from restricted or encrypted resources, sensitive context is leaked directly in the HTTP response payload. The attack flow involves a threat actor directly interacting with the exposed network service by invoking the /api/tag/getTag API route without supplying any authentication tokens or the document's publish password. Upon processing the request, the server serializes the comprehensive tag vocabulary—including metadata originating from confidential sources—and returns it to the client. This enables malicious actors to perform systematic reconnaissance, enumerating internal project names, confidential topics, and structural terminologies without authorization. The affected component is the tag management subsystem of SiYuan, impacting all deployments running versions prior to v3.7.4. Post-exploitation impact is characterized by unauthorized information exposure, facilitating further targeted attacks based on the leaked intelligence gathered from the internal document corpus."
}