CVE-2024-12450 is a critical (CVSS 9.8) set of flaws in the web_crawl function of infiniflow RAGFlow version 0.12.0, covering server-side request forgery, local file read and a path to remote code execution. The record says version 0.14.0 resolves them. Upgrade and keep the service off the public internet.
What's affected
- infiniflow/ragflow version 0.12.0, as named in the record.
The record says the issues are resolved in version 0.14.0. Confirm against the project release notes and the linked commit before you roll out.
How to check if you're exposed
Run the check below, or read the vendor or project notes linked in Sources.
docker images | grep -i ragflow
Compare the result to the affected range above.
What to do now
- Upgrade RAGFlow to 0.14.0 or later, using the project's release notes for any migration steps.
- Do not expose the RAGFlow web interface or API to the internet. Put it behind a VPN or an authenticating reverse proxy.
- Restrict outbound traffic from the RAGFlow host so it cannot reach internal admin interfaces or cloud metadata addresses.
- Review whether the container runs with reduced privileges. The record notes headless Chromium running with the sandbox turned off.
- Look up CVE-2024-12450 on its Gold page to see the current details and any fix information.
- Check your own dependency files with the free in-browser dependency scan. It runs in your browser and does not send the file to a server.
Why it matters
The record describes three linked problems in one function. First, the crawler does not filter URL parameters, so an attacker can point it at internal addresses and read the response through the generated PDF. Second, the file protocol is not restricted, which allows reading files on the server. Third, the headless Chromium build is old and runs in no-sandbox mode, which exposes the host to known browser engine bugs.
SSRF is the entry point, and the other two make it worse. A crawler that fetches pages for you is a feature that, left unconstrained, fetches pages for an attacker too. Fetchers should allowlist schemes and block private ranges, and browsers used for rendering should keep their sandbox on.
This CVE is a server-side request forgery flaw (CWE-918, Server-Side Request Forgery). Nothing in the record says it is exploited in the wild, and it is not marked as listed in CISA's Known Exploited Vulnerabilities catalog. That is a snapshot, so keep checking.
Server-side request forgery turns your web application into a proxy. The attacker supplies an address and your server makes the request, from inside your network, with your network permissions.
That matters because internal services often trust anything on the inside. Cloud metadata endpoints, admin panels bound to localhost and internal APIs may all answer a request that would never be allowed from the internet. The defenses are allowlisting the hosts a feature may contact, blocking private and link-local address ranges, and restricting outbound traffic from web servers at the network level.
Try it yourself
Click through it at your own pace, or open it full screen.
Quick check
Which version is fixed?
The record says 0.14.0. Verify in the project release notes.
Why is the old Chromium a problem?
An outdated headless browser carries known engine bugs, and running it without its sandbox removes a layer that would contain them.
Sources
Note: these pages could not be retrieved when this post was written, so statements rest on the CVE record text. Check the vendor pages for the latest version information.
Last reviewed: 19 October 2026.
