Executive brief
HyperChat is an AI agent platform that allows users to interact with various large language models. A security flaw in its AI Proxy Middleware allows an attacker to force the server to make unauthorized requests to internal or external websites. This could be used to scan internal networks, bypass firewalls, or access sensitive data stored on other internal systems that are not normally accessible from the internet.
Technical details
A server-side request forgery (SSRF) vulnerability exists in the AI Proxy Middleware of HyperChat (specifically in packages/core/src/http/aiProxyMiddleware.mts). The 'fetch' function processes the 'baseurl' request header without sufficient validation or allowlisting. An attacker can manipulate this header to coerce the server into making outbound HTTP requests to arbitrary destinations, including internal network services. The vulnerability is reachable via the /<password>/ai/... route. While a proof-of-concept is publicly available, no official patch has been released by the vendor at the time of reporting.
Affected products
- BigSweetPotatoStudio HyperChat up to 2.0.0-alpha.63
Timeline
- 2026-04-10: disclosed: Issue reported to the project maintainers via GitHub.
- 2026-04-28: advisory: Vulnerability details published.