GHSA-6MCF-G667-W3QV
Vulnerability from github – Published: 2026-09-03 22:50 – Updated: 2026-09-03 22:50CVE: This vulnerability corresponds to CVE-2026-72806.
Summary
FilterViewByPublishAccess, the filter renderAttributeView applies for Reader sessions drops rows using only the hidden/forbidden check and never checks the publish password. Its three sibling filters all check both tiers. As a result, a publish RoleReader (or the anonymous account when Publish.Auth.Enable is false) who has not entered a document's password still receives every database/attribute-view row bound to that password-protected document, the primary cell (title/ID) and all column values.
Details
FilterViewByPublishAccess (model/publish_access.go:290) drops rows on the hidden/forbidden tier only:
// Table (line 311), Gallery (348), Kanban (385), all identical:
if !CheckPathAccessableByPublishIgnore(bt.BoxID, bt.Path, publishIgnore) {
row = nil // hidden/forbidden dropped, but password NEVER checked
}
The three sibling filters all check both the hidden/forbidden tier and the password tier (password == "" || CheckPublishAuthCookie(...)):
- FilterBlockAttributeViewKeysByPublishAccess (line 412)
- FilterBlockInfoByPublishAccess (line 457)
- FilterContentByPublishAccess (line 474)
So the password (protected) tier is enforced everywhere except this AV/database-view filter. Table layout masks nothing; Gallery and Kanban mask only the cover, keeping the card and its non-cover values. Reachable via renderAttributeView, getAttributeViewKeys, and renderSnapshotAttributeView, all CheckAuth-only.
Proof of Concept
Reproduced on a live instance (publish mode on 6808, anonymous Reader, no password cookie), against a password-protected document with a database/AV row.
| Check | Path | Result |
|---|---|---|
| Control | getDoc(secretDoc) |
🔒 placeholder, body withheld password gate works normally |
| Test | renderAttributeView(AV) |
row leaked : blockID=…rk7jofz, title secret-db-row |
Differential (disable=true) |
same filter | 0 rows : hidden tier correctly enforced |
| Differential (password set) | same filter | 1 row : password tier bypassed |
Same filter, same document, same Reader: the hidden tier drops the row, the password tier leaks it isolating the omission.
Impact
An anonymous/Reader publish user who has not supplied a protected document's password receives all attribute-view/database rows bound to that document titles, block IDs, and column values defeating the publish-password control for database views. Confidentiality-only. The hidden/forbidden tier is unaffected (correctly enforced).
Suggested fix
Add the password check to the drop condition in all three layout branches (Table, Gallery, Kanban), mirroring the sibling filters:
if !CheckPathAccessableByPublishIgnore(...) ||
!(password == "" || CheckPublishAuthCookie(c, passwordID, password)) {
row = nil
}
{
"affected": [
{
"package": {
"ecosystem": "Go",
"name": "github.com/siyuan-note/siyuan/kernel"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "0.0.0-20260723040913-768427f20f13"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-72806"
],
"database_specific": {
"cwe_ids": [
"CWE-862"
],
"github_reviewed": true,
"github_reviewed_at": "2026-09-03T22:50:49Z",
"nvd_published_at": null,
"severity": "MODERATE"
},
"details": "**CVE:** This vulnerability corresponds to [CVE-2026-72806](https://nvd.nist.gov/vuln/detail/CVE-2026-72806).\n\n### Summary\n\n`FilterViewByPublishAccess`, the filter `renderAttributeView` applies for Reader sessions drops rows using only the hidden/forbidden check and never checks the publish password. Its three sibling filters all check both tiers. As a result, a publish `RoleReader` (or the anonymous account when `Publish.Auth.Enable` is `false`) who has not entered a document\u0027s password still receives every database/attribute-view row bound to that password-protected document, the primary cell (title/ID) and all column values.\n\n### Details\n\n`FilterViewByPublishAccess` (`model/publish_access.go:290`) drops rows on the hidden/forbidden tier only:\n```go\n// Table (line 311), Gallery (348), Kanban (385), all identical:\nif !CheckPathAccessableByPublishIgnore(bt.BoxID, bt.Path, publishIgnore) {\n row = nil // hidden/forbidden dropped, but password NEVER checked\n}\n```\n\nThe three sibling filters all check both the hidden/forbidden tier and the password tier (`password == \"\" || CheckPublishAuthCookie(...)`):\n- `FilterBlockAttributeViewKeysByPublishAccess` (line 412)\n- `FilterBlockInfoByPublishAccess` (line 457)\n- `FilterContentByPublishAccess` (line 474)\n\nSo the password (protected) tier is enforced everywhere except this AV/database-view filter. Table layout masks nothing; Gallery and Kanban mask only the cover, keeping the card and its non-cover values. Reachable via `renderAttributeView`, `getAttributeViewKeys`, and `renderSnapshotAttributeView`, all `CheckAuth`-only.\n\n### Proof of Concept\n\nReproduced on a live instance (publish mode on 6808, anonymous Reader, no password cookie), against a password-protected document with a database/AV row.\n\n| Check | Path | Result |\n|---|---|---|\n| Control | `getDoc(secretDoc)` | \ud83d\udd12 placeholder, body withheld password gate works normally |\n| Test | `renderAttributeView(AV)` | row leaked : `blockID=\u2026rk7jofz`, title `secret-db-row` |\n| Differential (`disable=true`) | same filter | 0 rows : hidden tier correctly enforced |\n| Differential (password set) | same filter | 1 row : password tier bypassed |\n\nSame filter, same document, same Reader: the hidden tier drops the row, the password tier leaks it isolating the omission.\n\n### Impact\n\nAn anonymous/Reader publish user who has not supplied a protected document\u0027s password receives all attribute-view/database rows bound to that document titles, block IDs, and column values defeating the publish-password control for database views. Confidentiality-only. The hidden/forbidden tier is unaffected (correctly enforced).\n\n### Suggested fix\n\nAdd the password check to the drop condition in all three layout branches (Table, Gallery, Kanban), mirroring the sibling filters:\n```go\nif !CheckPathAccessableByPublishIgnore(...) ||\n !(password == \"\" || CheckPublishAuthCookie(c, passwordID, password)) {\n row = nil\n}\n```",
"id": "GHSA-6mcf-g667-w3qv",
"modified": "2026-09-03T22:50:49Z",
"published": "2026-09-03T22:50:49Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/siyuan-note/siyuan/security/advisories/GHSA-6mcf-g667-w3qv"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-72806"
},
{
"type": "WEB",
"url": "https://github.com/siyuan-note/siyuan/commit/768427f20f13bbd8dc4effa8aa4e1d09a7741bf4"
},
{
"type": "PACKAGE",
"url": "https://github.com/siyuan-note/siyuan"
},
{
"type": "WEB",
"url": "https://www.vulncheck.com/advisories/siyuan-before-authentication-bypass-via-attribute-view"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:N/A:N",
"type": "CVSS_V3"
}
],
"summary": "SiYuan: Password (protected) tier omitted in the attribute-view/database publish filter: Reader receives rows of protected documents without the password (publish mode)"
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.