The gpt‑researcher project contains a cross‑site scripting flaw in the Report API endpoint in the backend/server/app.py component, active in all releases up to 3.4.3. The flaw allows remote attackers to inject malicious JavaScript that runs in the context of the victim’s browser when the report page is rendered, potentially stealing session cookies, hijacking accounts, or executing additional client‑side code. This issue is classified as CWE‑79 and involves code execution paths identified as CWE‑94. The impact is the loss of confidentiality and integrity of user data and the possibility of executing arbitrary client‑side code on users’ machines.

The vulnerability affects the gpt‑researcher application maintained by the user assafelovic, specifically the Report API logic in backend/server/app.py. All releases from the initial release through 3.4.3 are affected. The repository is hosted on GitHub and users can download the source or clone the project from https://github.com/assafelovic/gpt‑researcher.

The CVSS v3 score of 5.3 indicates a moderate risk, and the EPSS score is not available, so the exact likelihood cannot be quantified. The vulnerability is exploitable remotely, and published exploit scripts are available on GitHub and other vulnerability databases, making it a realistic threat. The issue is not currently listed in the CISA KEV catalog, but the existence of a publicly available exploit increases the urgency. Attackers can trigger the flaw by sending crafted requests to the Report API endpoint, and the resulting XSS can be used to steal session cookies or deliver further malicious payloads to unsuspecting users.