Description

Heym before 0.0.98 fails to apply SSRF egress guards to WebSocket Send and WebSocket Trigger nodes, allowing authenticated users to connect to internal services. Attackers can craft workflow nodes with arbitrary URLs and headers to reach internal services and read responses from the WebSocket Trigger node.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "5.4",
  "pubDate": "2026-09-01T16:17:35.763Z",
  "pubdate": "2026-09-01T16:17:35.763Z",
  "executiveSummary": "A Server-Side Request Forgery (SSRF) vulnerability exists in Heym versions prior to 0.0.98. The vulnerability originates from a failure to enforce egress filtering mechanisms on specific workflow automation nodes.\nSpecifically, the WebSocket Send and WebSocket Trigger nodes lack the necessary security guards to restrict outbound connections. This oversight allows authenticated users to manipulate workflow definitions to initiate requests against internal network resources that should otherwise be unreachable.\nBy crafting malicious workflow nodes with arbitrary URLs and custom headers, an attacker can coerce the application server into acting as a proxy. This capability allows the attacker to interact with internal services, bypass perimeter firewalls, and potentially exfiltrate sensitive data contained in responses from the WebSocket Trigger node.\nThe risk is significant for organizations relying on Heym for workflow automation, as it enables lateral movement and reconnaissance within internal infrastructure. Exploitation requires an authenticated session, though the impact is limited only by the network reachability of the underlying server hosting the Heym instance.",
  "technicalDetails": "The vulnerability resides within the internal architecture of Heym's workflow engine, specifically involving the processing logic for WebSocket communication nodes. Prior to version 0.0.98, the application failed to implement robust egress sanitization and validation for the WebSocket Send and WebSocket Trigger nodes.\nThe root cause is an insecure implementation of request handling, where the backend application fails to validate the destination URI against a defined allowlist or blocklist before executing the outbound WebSocket connection. Because the application processes these nodes with the privileges of the server-side process, it implicitly trusts the provided destination parameters.\nThe attack flow proceeds as follows: First, an authenticated attacker crafts a malicious workflow configuration containing a WebSocket Send or WebSocket Trigger node. Within the node configuration, the attacker specifies a target URL pointing to an internal service, such as a metadata endpoint (e.g., 169.254.169.254), an internal API, or a database management interface. The attacker may also inject arbitrary HTTP headers to bypass simple application-level authentication or to influence the behavior of the target service.\nWhen the workflow is executed or triggered, the Heym server initiates an outbound request to the attacker-defined internal address. Because the egress guards were absent, the request is dispatched without inspection. The target internal service processes the request and returns a response. If the payload involves the WebSocket Trigger node, the application facilitates the delivery of the response data back to the user-controlled workflow interface. This allows the attacker to read the output of the internal request, effectively turning the Heym server into a gateway to the internal network.\nThe vulnerability is limited to the server-side execution context; the attacker must have valid credentials to modify or trigger workflows. However, once authentication is established, the attacker can probe for internal services, perform port scanning, or interact with private network protocols, leading to full internal information disclosure or potential secondary exploitation of internal vulnerabilities that lack further authentication."
}