The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill directs users to download a toolkit from a GitHub repository not associated with a trusted organization or the skill's author.
Evidence: git clone https://github.com/Shantanu-U69/Spoof-Suite-Security.git in the Installation section of SKILL.md.
[REMOTE_CODE_EXECUTION]: The skill instructs the execution of a shell script with root privileges immediately after downloading it from the untrusted external repository.
Evidence: sudo ./install.sh in the Installation section of SKILL.md.
[PRIVILEGE_ESCALATION]: The instructions require extensive use of sudo for kernel module loading, system capability modification, and system service management. It specifically recommends disabling macOS System Integrity Protection (SIP), which significantly reduces the security posture of the host machine.
Evidence: sudo modprobe hardsp00f, sudo setcap cap_net_admin,cap_net_raw+eip /usr/bin/hardsp00fd, and the macOS: SIP must be disabled requirement in SKILL.md.
[PERSISTENCE]: The skill provides commands to enable a systemd service, ensuring the spoofing toolkit remains active across system reboots.
Evidence: sudo systemctl enable hardsp00fd in the Troubleshooting section.
[COMMAND_EXECUTION]: The skill relies on shell commands that modify the kernel, network stack, and system logs, often requiring elevated privileges.
Evidence: Use of modprobe, apt-get install, systemctl, and setcap commands throughout the file.
[INDIRECT_PROMPT_INJECTION]: The skill processes external YAML profile data and uses AI-generated content to influence system-level configurations (fingerprints, network strategies), which could be manipulated by malicious data sources.
Ingestion points: External YAML files loaded via hardsp00fd --profile and AI-generated profiles from ai_gen.generate_for_context.
Boundary markers: None identified in the skill instructions to delimit untrusted data.
Capability inventory: Kernel module modification, MAC address randomization, packet manipulation, and browser fingerprint injection.
Sanitization: No evidence of validation or sanitization for the generated or imported profiles.

spoof-suite-security-hardware