The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references and provides instructions for downloading multiple external security tools, including GoPhish, Evilginx2, the Social Engineer Toolkit (SET), and theHarvester. References point to well-known domains such as github.com and getgophish.com.
[COMMAND_EXECUTION]: SKILL.md contains numerous shell command examples for reconnaissance (theHarvester, dig), tool installation (apt install gophish), and framework configuration (evilginx2 setup).
[DATA_EXFILTRATION]: The management scripts (scripts/agent.py and scripts/process.py) utilize the requests library to communicate with a remote GoPhish API. These scripts explicitly disable SSL/TLS certificate verification using verify=False and urllib3.disable_warnings(), which increases the risk of man-in-the-middle attacks.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8).
Ingestion points: scripts/agent.py and scripts/process.py fetch campaign data, results, and target information from the external GoPhish API via the _get and get_campaign_results methods.
Boundary markers: No boundary markers or instructions to ignore embedded content are present in the processing logic.
Capability inventory: The scripts possess network request capabilities (requests.get/post) and file system write access (open().write()).
Sanitization: The skill lacks sanitization, escaping, or validation of the data retrieved from the API before it is printed to the console or written to markdown reports.

conducting-social-engineering-penetration-test