The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides a Bash script (recon.sh) for external attack surface enumeration that executes multiple specialized security binaries including subfinder, amass, dnsx, httpx, and naabu via the shell.
[INDIRECT_PROMPT_INJECTION]: The skill processes external inputs such as target domains and URLs in its reconnaissance and vulnerability testing scripts, presenting an attack surface for data-driven injections.
Ingestion points: The scripts accept a TARGET domain in recon.sh and a target_url in the SQLiTester Python class (SKILL.md).
Boundary markers: There are no explicit delimiters or instructions provided to the agent to differentiate between user instructions and potentially malicious content within the target data.
Capability inventory: The skill allows the agent to perform external network requests (via requests) and invoke local system commands (via the reconnaissance script).
Sanitization: The provided code samples do not perform validation or escaping of the target domain or URL strings before they are used in commands or network calls.

Penetration Tester