CVE-2025-65720

GPT Researcher v3.3.7 Command Injection Allowing Remote Code Execution

Published Jul 15, 2026 CVSS 3.1 · 9.8 Critical EPSS < 1% KEV · Not listed
Description
An issue in Open Source GPT Researcher v3.3.7 allows attackers to execute arbitrary commands on a victim system via user interaction with a crafted HTML page.

Analysis

AI-generated

Technical impact

Vulnerability in the open-source GPT Researcher application version 3.3.7 permits the injection and execution of arbitrary system commands. The flaw corresponds to CWE-77, a command injection weakness, and allows an attacker to run any command with the privileges of the GPT Researcher process when a victim interacts with a malicious HTML page. The high CVSS score of 9.8 reflects the severity and potential impact of this flaw.

Affected Systems

The affected product is the open‑source GPT Researcher platform, specifically release 3.3.7. User‑facing components that process HTML input include the web interface and any web server endpoints that serve user‑generated content. No alternative vendor or product names are listed.

Risk and Exploitability

The EPSS score indicates a very low‑but‑nonzero probability of exploitation, and the vulnerability is not currently listed in the CISA KEV catalog. The attack requires a user to interact with a crafted HTML page, which can be delivered via phishing or compromised websites. Based on the description, it is inferred that the attacker does not need authenticated or privileged access; the attacker only needs to entice a user to view the malicious content. Once triggered, the attacker can execute any system command with the privileges of the GPT Researcher application, potentially affecting confidentiality, integrity, and availability of the host system.

Remediation

No solution or workaround provided in the CVE record.

OpenCVE Recommended Actions

  • Upgrade to a patched release of GPT Researcher that fixes the command injection flaw.
  • Run the GPT Researcher process under a least‑privilege user or within a container that limits system command execution.
  • Implement input filtering or a web‑application firewall to sanitize or reject HTML content that contains suspicious command injection patterns before it reaches the vulnerable component.

Generated by OpenCVE AI on August 4, 2026 at 18:32 UTC.

Tracking

Sign in to view the affected projects.

Advisories

Source ID Title
Github GHSA Github GHSA GHSA-8j86-h8gg-797p GPT Researcher MCP STDIO configuration allows remote command execution
History
Type Values Removed Values Added
Title GPT Researcher v3.3.7 Command Injection Allowing Remote Code Execution
Type Values Removed Values Added
Title Command Injection in GPT Researcher Allows Arbitrary Code Execution
Type Values Removed Values Added
Title Command Injection in GPT Researcher Allows Arbitrary Code Execution
Type Values Removed Values Added
Weaknesses CWE-77
Metrics cvssV3_1

{'score': 9.8, 'vector': 'CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H'}

ssvc

{'options': {'Automatable': 'yes', 'Exploitation': 'none', 'Technical Impact': 'total'}, 'version': '2.0.3'}

Type Values Removed Values Added
First Time appeared Assafelovic
Assafelovic gpt-researcher
Vendors & Products Assafelovic
Assafelovic gpt-researcher
Type Values Removed Values Added
Description An issue in Open Source GPT Researcher v3.3.7 allows attackers to execute arbitrary commands on a victim system via user interaction with a crafted HTML page.
References
Weaknesses
  • CWE-77

    Improper Neutralization of Special Elements used in a Command ('Command Injection')

Sources

Status
PUBLISHED
Assigner
mitre
Published
Updated
Jul 16, 2026, 6:12 PM
Reserved
Nov 18, 2025, 12:00 AM
Status
Awaiting Analysis
Published
Jul 15, 2026, 10:16 PM
Modified
Oct 6, 2026, 10:10 PM