LobeChat, a Node.js-based conversational AI application, contains a regular expression denial-of-service flaw before version 2.2.10-canary.15. An authenticated attacker can embed a catastrophic-backtracking regular-expression pattern in a GitHub repository URL path. When the application builds a dynamic regex from the unvalidated basePath, the pattern causes the regex engine to perform exponential backtracking, blocking the Node.js event loop. This results in a temporary denial of service that impacts all concurrent users for several seconds per request.

All releases of LobeChat from lobehub built before 2.2.10-canary.15 are vulnerable. The issue is triggered during the import of a GitHub skill when an attacker supplies a malicious repository URL. LobeChat is a Node.js-based conversational AI application developed by lobehub.

With a CVSS score of 7.1 the issue is rated high severity. The EPSS score of less than 1% indicates a low likelihood of real-world exploitation at present, and the vulnerability is not listed in the CISA KEV catalog. Exploitation requires the attacker to be authenticated to LobeChat and to trigger a skill import with a specially crafted GitHub URL that contains a backtracking pattern. The attack is performed synchronously and can be repeated to sustain a denial of service, though the description does not explicitly confirm repeatability.