The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.run and os.system via xargs to execute standard networking and security tools (e.g., curl, dig, nmap, subfinder, gau). These executions are core to the skill's purpose for security reconnaissance.
Evidence: Found in recon_pipeline.sh and findings_gen.py. The findings_gen.py script uses curl to perform live WordPress detection.
Mitigation: The skill implements a strict regex-based domain validator (DOMAIN_RE) across all entry points (recon_pipeline.sh, findings_gen.py, build_xlsx.py) to prevent shell metacharacter injection and directory traversal attacks.
[EXTERNAL_DOWNLOADS]: The skill downloads data from common OSINT and security service APIs.
Evidence: Fetches data from rdap.org, certspotter.com, crt.sh, and hudsonrock.com (a well-known breach intelligence provider). These are standard sources for external reconnaissance and are used in a read-only capacity to populate the engagement evidence tree.

osint-autopilot