CVE-2026-91943

SSRF in Crawl4AI PDFContentScrapingStrategy allows internal network access via malicious PDF URLs.

• CVSS 8.3
• 918
• Input Validation and Sanitization
• Remote

Crawl4AI before 0.9.3 contains a server-side request forgery vulnerability in PDFContentScrapingStrategy where _get_pdf_path() re-downloads targets with Python requests without egress validation. Authenticated attackers can supply URLs that redirect to internal addresses or use DNS rebinding to access internal services, exfiltrating responses through PDF text extraction in crawl results.

CVE

CVE-2026-91943

CWE

918

CVSS base score

8.3

Published

2026-09-15

OWASP

A10 Server-Side Request Forgery

Orthogonal defect classification

Checking

Code defect classification

Missing Check

Category

Input Validation and Sanitization

Subcategory

Insecure Direct Object References (IDOR)

Accessibility scope

Remote

Impact

Information Disclosure

Affected component

Python

Fixed by upgrading

Yes

Solution

Upgrade Crawl4AI to version 0.9.3 or later.

import requests
from urllib.parse import urljoin

class PDFContentScrapingStrategy:
    def __init__(self, base_url):
        self.base_url = base_url

    def _get_pdf_path(self, pdf_url):
        # VULNERABLE: no validation of pdf_url, allows SSRF
        response = requests.get(pdf_url, timeout=10)
        response.raise_for_status()
        return response.content

    def scrape(self, pdf_url):
        pdf_bytes = self._get_pdf_path(pdf_url)
        # pretend to extract text from PDF bytes
        return pdf_bytes[:100]

import requests
from urllib.parse import urlparse, urljoin

class PDFContentScrapingStrategy:
    def __init__(self, base_url):
        self.base_url = base_url

    def _get_pdf_path(self, pdf_url):
        # FIX: validate scheme and netloc to prevent SSRF
        parsed = urlparse(pdf_url)
        if parsed.scheme not in ("http", "https"):
            raise ValueError("Unsupported URL scheme")
        if not parsed.netloc:
            raise ValueError("URL must have a network location")
        response = requests.get(pdf_url, timeout=10)
        response.raise_for_status()
        return response.content

    def scrape(self, pdf_url):
        pdf_bytes = self._get_pdf_path(pdf_url)
        # pretend to extract text from PDF bytes
        return pdf_bytes[:100]

__VAITP_MODEL_REFUSED__

@misc{vaitp:cve202691943,
  title        = {{SSRF in Crawl4AI PDFContentScrapingStrategy allows internal network access via malicious PDF URLs.}},
  author       = {Bogaerts, Fr\'ed\'eric and Ivaki, Naghmeh and Fonseca, Jos\'e},
  year         = {2026},
  note         = {VAITP Python Vulnerability Dataset, entry CVE-2026-91943},
  howpublished = {\url{https://netpack.pt/vaitp/vulnerability/CVE-2026-91943/}}
}