SQL Injection in SiYuan Backlink Search Query Processing

SiYuan versions <= v3.7.2 contain a SQL injection vulnerability in the backlink/mention search query (kernel/model/backlink.go), which concatenates stored block metadata (title, name, alias, anchor text) and the client-supplied keyword into a SQL MATCH/search statement while escaping only the double-quote character and not the single quote. A single quote in the client keyword (first-order, reachable by an anonymous or RoleReader user on the publish surface) or in stored document metadata (second-order) breaks out of the string literal. Because the query runs on the main read-write siyuan.db handle via a statement-stacking-capable driver, an attacker can execute arbitrary SQL, enabling cross-notebook read and write. Fixed in v3.7.4.

Casky was already ahead

This CVE exploits attack patterns that Casky's 322matched skills already investigate — long before this vulnerability was disclosed. Claude's reasoning model maps these techniques to MITRE ATT&CK, so practitioners who ran these skills have already seen the threat behaviour in their findings.

Analysis

SiYuan versions 3.7.2 and earlier contain a critical SQL injection vulnerability in the backlink and mention search functionality. The vulnerability stems from improper input sanitization in kernel/model/backlink.go, where user-supplied keywords and stored block metadata (titles, names, aliases, anchor text) are directly concatenated into SQL MATCH statements with only double-quote escaping applied. Single quotes bypass this insufficient filtering, allowing attackers to manipulate SQL queries. The vulnerability is particularly severe because it's reachable by unauthenticated users and those with minimal RoleReader permissions on published surfaces, combined with a CVSS 10.0 critical rating. Any SiYuan deployment with versions at or below v3.7.2 faces immediate risk of data exfiltration, integrity compromise, or denial of service through malicious search queries.

Casky's 322 mapped security skills leverage Claude AI's extended reasoning to detect SQL injection attack patterns aligned with MITRE ATT&CK techniques TA0001 (Initial Access) and TA0043 (Reconnaissance). Practitioners would observe findings identifying: (1) suspicious single-quote characters embedded in backlink search parameters that attempt SQL syntax breaking, (2) anomalous UNION SELECT or comment-based SQL injection payloads in stored metadata fields, (3) timing-based blind SQL injection attempts indicating database query manipulation, and (4) unusual database query patterns during reconnaissance phases where attackers probe for injectable parameters. The platform's continuous analysis surfaces both first-order injection attempts from client keywords and second-order injection risks from previously-stored malicious metadata, enabling security teams to identify exploitation before successful compromise occurs.

Live Threat IntelligenceComing soon

Casky Risk Score

—

EPSS Probability

—

Shodan Exposure

—

GreyNoise Signal

—

Composite risk scoring from EPSS, CISA KEV, Shodan, and GreyNoise — 21 security APIs correlated into a single Casky Risk Score. Coming in Casky Pro. Join early access →

322 Casky skills cover this attack pattern

These skills use Claude AI's reasoning model to surface findings in the same attack categories as CVE-2026-72811.

Abuse of Public-Facing API: Mobile API Abuse

mobile security · low

Run

abusing-dpapi-for-credential-access

red teaming · high

Run

abusing-shadow-credentials-for-privesc

red teaming · high

Run

MITRE ATT&CK Techniques

TA0001 TA0043

Investigate the attack patterns behind CVE-2026-72811

Casky has 322 skills that investigate the attack patterns behind CVE-2026-72811. Run one and get CVSS-scored findings in 3 minutes.

Run the skill that detects this →

X Instagram LinkedIn