The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references and provides instructions to download code from two unverified GitHub repositories: https://github.com/jeroennijhof/vncpwd and https://github.com/trinitronx/vncpasswd.py. These sources are not from verified or well-known organizations.
[REMOTE_CODE_EXECUTION]: The skill instructs the agent/user to compile and run code retrieved from unverified external sources, specifically using make and executing the resulting binary ./vncpwd, or running a downloaded Python script vncpasswd.py.
[COMMAND_EXECUTION]: The skill executes multiple security-oriented CLI tools (Nmap, Hydra, Medusa, Ncrack, Metasploit) that perform active network operations and authentication testing against target systems.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The agent is instructed to read and analyze outputs from network discovery tools (vnc-info, vnc-title) and local system files (~/.vnc/passwd, ultravnc.ini).
Boundary markers: Absent. The instructions do not specify the use of delimiters or provide warnings to the agent to ignore potentially malicious content embedded in the processed data.
Capability inventory: The skill utilizes significant capabilities including network scanning, credential testing, and local file access across multiple scripts.
Sanitization: Absent. There is no evidence of validation or sanitization applied to the data retrieved from network services or configuration files before it is processed by the agent.

vnc-pentesting