The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides a wrapper script ht_run.py which allows the agent to execute arbitrary shell commands on the host system or within Docker containers via the --command flag.
[PRIVILEGE_ESCALATION]: The skill is designed to automatically escalate privileges by attempting to run commands with sudo -n when it encounters permission errors. It also supports a --privileged flag for Docker execution, which grants the container access to all host devices.
[EXTERNAL_DOWNLOADS]: The skill pulls and executes external Docker images from various repositories, including instrumentisto/nmap, projectdiscovery/nuclei, and kalilinux/kali-rolling. It also provides an --install mechanism to download and set up additional tools from external sources.
[DYNAMIC_EXECUTION]: The ht_run.py script dynamically constructs and executes command strings at runtime based on tool identifiers and arguments provided by the agent or user.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources such as domain records, website content, and security scan results, which could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: The agent processes output from tools like sherlock, maigret, nuclei, and httpx (e.g., SKILL.md workflows).
Boundary markers: No specific boundary markers or instructions to ignore embedded commands are present in the workflows or handling instructions.
Capability inventory: The skill has powerful capabilities including raw network socket access, filesystem writes, and shell command execution via the ht_run.py script.
Sanitization: The instructions do not define any sanitization or validation procedures for tool outputs before they are processed by the agent.

pentest