The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The installation process requires cloning a repository from an unverified GitHub user (Shantanu-U69) and executing build commands, kernel module installations (modprobe), and Python setup scripts with system-level privileges.
[PRIVILEGE_ESCALATION]: Almost all described operations require sudo or Administrator privileges. The skill installs kernel modules, Windows drivers, modifies machine-level environment variables, and assigns powerful network capabilities (cap_net_admin, cap_net_raw) to binaries.
[EXTERNAL_DOWNLOADS]: The Windows installation instructions direct users to download and execute binary archives and driver installation scripts from an unverified GitHub Pages site (shantanu-u69.github.io).
[INDIRECT_PROMPT_INJECTION]: The skill includes an AI integration component (LLMProfileGenerator) that processes a context dictionary to generate hardware profiles. This creates an attack surface for indirect prompt injection if the ingested context contains untrusted data.
Ingestion points: The context parameter in LLMProfileGenerator.generate_profile and local YAML profile paths.
Boundary markers: None identified in the provided instructions or examples.
Capability inventory: Kernel-level network spoofing (ARP, DNS, SNI), MAC address manipulation, file system writes for profile management, and driver installation.
Sanitization: While a ProfileValidator is mentioned, its efficacy is limited to schema validation rather than sanitizing natural language inputs.
[COMMAND_EXECUTION]: The skill relies heavily on the execution of shell commands and scripts to manage daemons, interfaces, and system configurations under elevated privileges.

spoof-suite-hardware-identity