CVE-2026-117569
HighCVSS 7.8Exploitation Probability (EPSS)
Elevated risk52th percentile - higher than 52% of all known CVEs
Summary
Kali-GPT through commit a608c74 contains an OS command injection vulnerability in execute_tool() in kali_gpt/tools/mcp_server.py. MCP clients can run commands via unsanitized scanner parameters such as target, ports, domain, wordlist, or url, executed via subprocess.run with shell=True.
Risk Assessment
Attackers or prompt-injected agents can execute arbitrary commands as the MCP server user, potentially leading to full system compromise.
Recommendation
Update Kali-GPT to a patched version or restrict access to the MCP server. Consider running the server in an isolated environment with minimal privileges.
Related vulnerabilities
- CVE-2026-121802Critical
@better-auth/expo before 1.7.16 signs a caller-supplied state value without verifying that it was issued by the server. This allows account-takeover outcomes, such as unauthenticated changes of any account's email address.
- CVE-2026-121615Critical
Linksys E5600 1.1.0.26 contains an improper authentication vulnerability in /cgi-bin/mesh_login.cgi (Login Handler). The attack can be launched remotely and the exploit has been publicly disclosed. The vendor did not respond and the product is no longer supported.
- CVE-2026-121609Critical
A vulnerability was identified in the Qihoo 360 Whole Home Router M5 version 4.0.17.1566 within the boot script /etc/init.d/mpt (Home function). The issue involves hard-coded credentials. The exploit is publicly available and remote exploitation is possible.
- CVE-2026-121607Critical
A vulnerability was found in Qihoo 360 D603-2, D600-2, D606-6 and D706 up to version 6.4.9 in the testspeed_command function of /fs2/bin/iscm. Manipulating the host argument results in OS command injection. The attack can be initiated remotely and the exploit has been made public.
- CVE-2026-121606Critical
A vulnerability was found in Qihoo 360 devices (models D621-2, D603-2, D600-2, D619, D606-6 and D706 up to 6.4.9) in the verify_sign function of fs2/bin/iscm responsible for firmware signature verification. Manipulating the sign argument leads to improper verification of cryptographic signatures. The attack can be launched remotely and the exploit has been publicly disclosed.
- CVE-2026-121600Critical
A vulnerability was found in TOTOLINK A950RG version 4.1.2cu.5268_B20221104 in the setMacFilterRules function of /cgi-bin/cstecgi.cgi handling MAC filtering. Manipulating the addEffect argument causes a stack-based buffer overflow. The attack can be initiated remotely and the exploit is publicly available.
- CVE-2026-121599Critical
A vulnerability was detected in TOTOLINK A950RG version 4.1.2cu.5268_B20221104 in the setWiFiMultipleConfig/delWiFiMultipleConfig functions of /cgi-bin/cstecgi.cgi handling Multi-SSID configuration. Manipulating the wepkey2/wepkey3 arguments results in a stack-based buffer overflow. The attack can be launched remotely and the exploit is now public.
- CVE-2026-121598Critical
A security vulnerability was detected in TOTOLINK A950RG version 4.1.2cu.5268_B20221104 in the strcpy function of /cgi-bin/cstecgi.cgi handling basic wireless configuration. Manipulating the wepkey argument leads to a stack-based buffer overflow. The attack can be initiated remotely and the exploit has been publicly disclosed.
- CVE-2026-92190Critical
Quest Foglight for Databases contains a deserialization of untrusted data vulnerability in the baseRmiServiceExporter component. An unauthenticated remote attacker can execute arbitrary code in the context of SYSTEM.
- CVE-2026-92189Critical
In Quest Foglight for Databases, the writeTopologySectionToFile method does not validate a user-supplied path before using it in file operations. An unauthenticated remote attacker can execute arbitrary code in the context of SYSTEM.
Original NVD description (English source)
Kali-GPT through commit a608c74 contains an OS command injection vulnerability in execute_tool() in kali_gpt/tools/mcp_server.py that allows MCP clients to run commands via unsanitized scanner parameters. Attackers or prompt-injected agents can embed shell metacharacters in parameters like target, ports, domain, wordlist, or url, executed via subprocess.run with shell=True as the MCP server user.
Vulnerability data from NVD (NIST) · CISA KEV · EPSS

