The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to invoke nmap and masscan for network reconnaissance. These operations include SYN scans, service version detection, OS fingerprinting, and NSE script execution, which are standard for the stated pentesting and bug bounty use cases.
[INDIRECT_PROMPT_INJECTION]: The skill ingests target data from external files such as scope.txt and exclude.txt.
Ingestion points: Hostnames and IP ranges are sourced from scope.txt and exclude.txt (SKILL.md).
Boundary markers: The instructions explicitly reference a tradecraft-scope-roe process to confirm targets are in-scope before execution.
Capability inventory: The skill performs network-level scanning and service fingerprinting via subprocess calls to nmap and masscan.
Sanitization: Standard CLI validation for IP addresses and hostnames is performed by the underlying security tools.

tools-nmap