CVE-2026-69084

CRITICAL

SiYuan before v3.7.3 SQL Injection via searchEmbedBlock

Title source: CVE List
Search associated code STIX 2.1 vulnerability STIX 2.1 with PoCs

Record summary

CVE-2026-69084 has a selected CVSS score of 9.9 (critical); EIP currently links 1 repository PoC.

Description

SiYuan versions <= v3.7.2 expose the /api/search/searchEmbedBlock endpoint, which passes a client-supplied SQL statement verbatim to the main read-write siyuan.db handle with no single-statement, read-only, or admin restrictions. The endpoint is gated only by CheckAuth, making it reachable by the publish RoleReader token and by anonymous users when publish authentication is disabled. Because the underlying driver executes stacked statements, an attacker can read and modify content across all opened cleartext notebooks (encrypted per-box notebooks are excluded). Fixed in v3.7.3.

Description source: CVE List

Exploitation context

Available material

Repository PoCs
1

CISA SSVC decision

ExploitationPoC
AutomatableYes
Technical impactTotal

CISA Coordinator · SSVC 2.0.3 · Evaluated Aug 3, 2026 · Source: CVE List

Affected products and versions

1
ProductSourceVersion range

siyuan

Browse siyuan-note / siyuan

Default status: unaffected

CVE ListBefore 3.7.3affected
3.7.3unaffected

Proofs of concept

1

Repository PoCs

GitHubBoreas37/CVE-2026-69084-PoCRepository PoCby Boreas371Exploit3 files

11.1 KiB

GitHub
PoC details
Analysisdeepseek-v4-pro:cloud ·

Technical assessment

Python script that exploits CVE-2026-69084 by sending arbitrary SQL statements to the /api/search/searchEmbedBlock endpoint of SiYuan. It includes a --proof mode that creates a table and inserts a row to demonstrate write access, and a --check mode for non-destructive probing.

Backdoor review

No backdoor observed in reviewed code

The PoC script and README demonstrate the documented CVE-2026-69084 SQL injection exploit against SiYuan. The Python code performs only the described actions: session login, sending attacker-supplied SQL to the vulnerable endpoint, and displaying results. No hidden, deceptive, or unrelated behavior is present.

ClassificationExploit
Model confidence98%
AuthenticationNot required
Languagespython
Target softwaresiyuan
Attack typessql injection
Evidence & reasoningClassification basis · observed behavior · safety review
Technical evidence

Classification basis and observed behavior

Classification basis

The script is designed to exercise the vulnerability by sending arbitrary SQL statements to the vulnerable endpoint, including destructive write operations (CREATE TABLE, INSERT) in --proof mode. This constitutes exploit code.

CVE-2026-69084.py:85-109CVE-2026-69084.py:118-127

Requirements

  • Target SiYuan instance <= v3.7.2 with the /api/search/searchEmbedBlock endpoint reachable.CVE-2026-69084.py:6
  • If publish authentication is enabled, a valid auth code is required; otherwise, the endpoint is accessible anonymously.CVE-2026-69084.py:14-16

Observed behavior

  • Performs a session login via /api/system/loginAuth if an auth code is provided.CVE-2026-69084.py:74-82
  • Sends a POST request to /api/search/searchEmbedBlock with a client-supplied SQL statement in the 'stmt' parameter.CVE-2026-69084.py:85-95
  • In --proof mode, executes CREATE TABLE and INSERT statements to write to the database.CVE-2026-69084.py:118-127
  • In --check mode, executes a non-destructive SELECT sqlite_version() query.CVE-2026-69084.py:112-115
Safety-review evidence

Behaviors behind the backdoor verdict

Observables

Url
http://127.0.0.1:6806Target host used in usage examples and verification output; standard for local testing.CVE-2026-69084.py:31README.md:52
Sql Statement
CREATE TABLE IF NOT EXISTS pwn_69084 (id INTEGER)Proof-of-concept SQL payload that creates a table to demonstrate write capability; matches the documented exploit.CVE-2026-69084.py:121
Sql Statement
INSERT INTO pwn_69084 VALUES (31337)Proof-of-concept SQL payload that inserts a row to demonstrate write capability; matches the documented exploit.CVE-2026-69084.py:122
Review boundaries

What the analysis did not establish

  • One file (LICENSE) was omitted from the evidence; it is unlikely to affect classification.
  • The evidence does not include the execution output or verification that the script successfully exploited a live target.
  • One file (LICENSE, 1036 bytes) was classified as binary and not analyzed for text content; it is a standard license file and unlikely to contain executable behavior.
  • The review is limited to the supplied text evidence; no dynamic analysis was performed.
Model interpretation

This review is limited to the supplied PoC code and context. It does not assert that the code works or is safe to execute.

References

3