Skip to content
Fencer Intel
high 8.1

GHSA-g485-8j3v-p6x8 json

@tdurieux/anonymous_github Vulnerable to XSS via Unsanitized GitHub Repository Content Rendering in Anonymous GitHub Origin

published 6mo ago modified 6mo ago

cvss
8.1
version 3.1, GitHub
cisa kev
Not listed
not in CISA's catalog
epss
—
no score, EPSS covers CVEs only
public exploits
0
none known

Summary

### Summary

Anonymous GitHub fetches repository content (e.g., markdown files) from GitHub's API and renders it without sanitization. On the client side, markdown is parsed with `marked` (with `sanitize: false`) and injected into the DOM via `$sce.trustAsHtml()` + `ng-bind-html`, bypassing AngularJS's built-in XSS protection. An attacker can craft a malicious GitHub repository whose README executes arbitrary JavaScript in the Anonymous GitHub origin.

### Details

#### README fetched from GitHub API

The server fetches the README via GitHub's REST API and stores the raw markdown in MongoDB:

```typescript // https://github.com/tdurieux/anonymous_github/blob/b2d77faa6c6f35ad9ae6ed46e3a3fa4681ac84c2/src/core/source/GitHubRepository.ts#L162-L174 const ghRes = await oct.repos.getReadme({ owner: this.owner, repo: this.repo, ref: selected?.commit, }); const readme = Buffer.from( ghRes.data.content, ghRes.data.encoding as BufferEncoding ).toString("utf-8"); selected.readme = readme; await model.save(); ```

It is then served to the client with no sanitization:

```typescript // https://github.com/tdurieux/anonymous_github/blob/b2d77faa6c6f35ad9ae6ed46e3a3fa4681ac84c2/src/server/routes/repository-private.ts#L254-L260 return res.send( await repo.readme({ accessToken: token, force: req.query.force == "1", branch: req.query.branch as string, }) ); ```

#### Client-side rendering via `$sce.trustAsHtml()` + `ng-bind-html`

The client fetches the raw README, parses it with `renderMD()` (which uses `marked` with `sanitize: false`), then bypasses AngularJS sanitization:

```javascript // https://github.com/tdurieux/anonymous_github/blob/b2d77faa6c6f35ad9ae6ed46e3a3fa4681ac84c2/public/script/app.js#L1219-L1226 const res = await $http.get(`/api/repo/${o.owner}/${o.repo}/readme`, { params: { force: force === true ? "1" : "0", branch: $scope.source.branch }, }); $scope.readme = res.data; ```

```javascript // https://github.com/tdurieux/anonymous_github/blob/b2d77faa6c6f35ad9ae6ed46e3a3fa4681ac84c2/public/script/app.js#L1339-L1343 const html = renderMD( $scope.anonymize_readme, `https://github.com/${o.owner}/${o.repo}/raw/${$scope.source.branch}/` ); $scope.html_readme = $sce.trustAsHtml(html); // sink: bypasses Angular XSS protection ```

The `renderMD()` function explicitly disables sanitization:

```javascript // https://github.com/tdurieux/anonymous_github/blob/b2d77faa6c6f35ad9ae6ed46e3a3fa4681ac84c2/public/script/utils.js#L165-L176 marked.setOptions({ sanitize: false, // HTML in markdown is preserved as-is // ... }); return marked.parse(md, { renderer }); ```

The resulting HTML is bound to the DOM via `ng-bind-html`, which trusts the string marked by `$sce.trustAsHtml()` and inserts it as innerHTML.

### Impact

1. **Stored XSS**: Any malicious GitHub repository can execute JavaScript in the Anonymous GitHub origin when a user anonymizes it or views its content 2. **Account Takeover**: Steal authentication tokens and session cookies 3. **Data Exfiltration**: Access other users' anonymization configurations and private repository data via `/api/user` and `/api/repo/list`

### Proof of Concept

![poc-xss-anonymous-github](https://github.com/user-attachments/assets/c1bf3ed9-4e1e-4c8e-87f0-782a8d5f6ead)

1. Create a GitHub repository with a malicious `README.md`:

```markdown # Innocent README <img src=x onerror="alert(document.domain)"> ```

2. On Anonymous GitHub, enter the malicious repository URL to anonymize it 3. The XSS executes immediately when the README preview is rendered on the anonymize page

### Remediation

1. Sanitize markdown output with DOMPurify before rendering (the dependency already exists but is unused) 2. Serve HTML files with `Content-Disposition: attachment` or in a sandboxed iframe on a separate origin 3. Replace `$sce.trustAsHtml()` with proper `ngSanitize` usage 4. HTML-escape filenames and paths in directory listing templates 5. Add Content Security Policy headers

### Credits

Zhengyu Liu, Jingcheng Yang

Patch / fix

  • npm @tdurieux/anonymous_github 2.3.0

Is GHSA-g485-8j3v-p6x8 in your dependencies? Fencer shows which of your repositories are affected as soon as an advisory lands, and opens the fix as a pull request.

Check with Fencer

CVSS scores

assigned by version score vector
GitHub 3.1 8.1 CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N

Affected packages

ecosystem package affected fixed in
npm @tdurieux/anonymous_github
>=2.2.0, <2.3.0
2.3.0

Classification

Weakness
  • CWE-79 Improper Neutralization of Input During Web Page Generation
  • CWE-80 Improper Neutralization of Script-Related HTML Tags in a Web Page

References

Sources

Fencer Intel keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.