The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides numerous instructions for the agent to execute shell commands directly on the host system, including file system operations and network utility calls.
[COMMAND_EXECUTION]: Multiple workflows explicitly require the use of sudo to perform SYN scans, UDP scans, and OS fingerprinting, granting the agent elevated system-level permissions.
[COMMAND_EXECUTION]: The implementation workflow relies on capturing command output from files (e.g., .nmap results) and interpolating those variables directly into subsequent shell commands, which is a dangerous pattern that could allow for command injection if the source data is manipulated.
[PROMPT_INJECTION]: The skill contains an indirect prompt injection surface through its automated parsing of network scan results.
Ingestion points: Data is read from nmap-portscan.nmap and nmap-portscan.gnmap during the Phase 2 workflow in SKILL.md.
Boundary markers: There are no explicit delimiters or instructions to ignore embedded instructions within the processed scan data.
Capability inventory: The agent is granted the ability to execute mkdir, nmap, and several shell utilities including grep, sed, cut, and tr.
Sanitization: While the skill uses basic regex for port identification, it lacks robust sanitization to prevent adversarial payloads in service names or version strings from affecting subsequent command execution.

Nmap Scan