/

CVE-2026-91079

Severity:

HIGH RISK

(8.5)

Server-Side Request Forgery (CWE-918) in Huly Print Service

CVE-2026-91079

Deferred

Vulnerability type

Category:

Server-Side Request Forgery

CWE:

CWE-918

Impact:

Internal Network and Metadata Service Access

Severity:

HIGH RISK

(8.5)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

Huly Platform contains a server-side request forgery vulnerability in its print service because supplied URLs are not restricted by a hostname allowlist. An authenticated workspace member may cause the service to render attacker-selected internal or external resources through Puppeteer and receive the results as downloadable PDFs or images.

Affected context

The vulnerability occurs in Huly Platform deployments where:

  • Authenticated workspace members can access the print endpoint
  • The endpoint accepts a user-controlled URL for rendering
  • The print service uses Puppeteer to retrieve and render that URL
  • Hostname, destination, or network-boundary validation is missing or insufficient

Risk depends on the network access available to the print-service process and the sensitivity of reachable internal resources.

Exploit preconditions

  • Attack surface: The Huly print service endpoint
  • Required access: An authenticated workspace-member account with permission to invoke the print functionality
  • Trigger point: Submission of an attacker-controlled URL for PDF or image rendering
  • Why exploitation is possible: The service makes outbound requests on behalf of the requester without reliably restricting destinations through hostname or network allowlists

Why this happens

The underlying flaw is server-side request forgery caused by treating a user-supplied rendering URL as an unrestricted destination. The print service's browser automation process can resolve and request hosts that the requester may not be able to access directly, including internal services and metadata endpoints.

This maps to CWE-918: Server-Side Request Forgery, where a server performs network requests influenced by an external actor without adequate destination validation or network-boundary controls.

Potential impact

  • An attacker may retrieve content from internal network hosts
  • The service could expose cloud or infrastructure metadata
  • Internal administrative interfaces may be probed or accessed
  • Sensitive responses may be returned as downloadable documents or images
  • The vulnerability could support further compromise if reachable services contain exploitable functions

General mitigation guidance

Mitigation generally focuses on making outbound rendering destinations explicit and minimizing the print service's network privileges.

Common defensive patterns include:

  • Enforcing strict hostname and destination allowlists
  • Resolving hostnames and validating resulting addresses before making requests
  • Blocking loopback, link-local, private, reserved, and cloud-metadata network ranges
  • Rechecking redirects so a permitted URL cannot reach a prohibited destination
  • Restricting outbound network access with firewall or workload policies
  • Separating rendering from sensitive internal networks
  • Running the browser-rendering process with least-privilege identity and filesystem access
  • Returning only the intended rendered representation rather than unrestricted response data

How teams detect issues like this

Detection approaches commonly include:

  • Static analysis for server-side HTTP, browser, or Puppeteer navigation using user-controlled URLs
  • Code review checks for missing hostname, IP-range, redirect, and DNS-rebinding validation
  • Tests covering loopback, private-address, link-local, metadata-service, and redirect destinations
  • Network monitoring for unexpected connections from the print-service workload
  • Egress policies that alert on or block requests to internal and infrastructure-management ranges
  • Application logging of requesting identity, submitted destination, resolved address, and redirect chain

Automated security review tools may help identify URL-to-network-request data flows, but deployment-level egress controls remain important.

Sources

SourceLink
github.comgithub.com/hcengineering/platform
github.comgithub.com/hcengineering/platform/blob/v0.7.426/services/print/pod-print/src/config.ts
github.comgithub.com/hcengineering/platform/blob/v0.7.426/services/print/pod-print/src/server.ts
github.comgithub.com/hcengineering/platform/issues/10908
vulncheck.comvulncheck.com/advisories/huly-platform-through-0.7.426-ssrf-via-print-service

Frequently asked

questions

It may be exploitable through the Huly print endpoint by an authenticated workspace member, provided that the print service can reach the selected internal or external destination.

The described attack requires authenticated workspace-member access and the ability to submit a URL to the print functionality. The resulting impact also depends on the print service's network connectivity and permissions.

Restrict rendering to explicitly approved destinations, validate resolved addresses and redirects, and prevent the rendering workload from reaching sensitive network ranges.

Treat all user-controlled URLs as untrusted, apply centralized SSRF protections, enforce outbound network policy, test DNS and redirect edge cases, and run rendering services with minimal privileges.

Find out what's already
exploitable in your codebase.