The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/agent.py uses the python-nmap library to execute the nmap command-line tool. This is the intended primary function for network discovery and service enumeration.
[PROMPT_INJECTION]: The skill processes untrusted data from network service banners and Nmap Scripting Engine (NSE) outputs, creating a surface for indirect prompt injection.
Ingestion points: Service banners, OS fingerprinting results, and NSE script outputs are read into the agent context via scripts/agent.py during the service version and vulnerability scan steps.
Boundary markers: The skill does not implement boundary markers or instructions to ignore instructions found within network responses.
Capability inventory: The skill executes system commands through the nmap binary and records results to the local filesystem.
Sanitization: No validation or sanitization is performed on the data returned by network services before it is included in JSON and CSV output reports.

scanning-network-with-nmap-advanced