The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires cloning a repository from an external source (https://github.com/0x4m4/hexstrike-ai.git) that is not associated with a verified or trusted vendor. This repository contains the core logic for the MCP server and tool wrappers.
[COMMAND_EXECUTION]: The installation process involves executing shell commands to install dependencies from a requirements file and running a Python script as an MCP server. This provides a direct path for executing code from the downloaded third-party source on the host system.
[PROMPT_INJECTION]: The skill's architecture presents a significant surface for indirect prompt injection because it ingests and processes data from over 150 external security tools. Malicious data encountered during a scan (e.g., in a web response or service banner) could potentially influence the AI agent's decision-making logic or exploit generation.
Ingestion points: Data returned by any of the 150+ integrated tools such as Nmap, Burp Suite, or SQLMap as described in SKILL.md.
Boundary markers: Absent; there are no instructions to the agent to treat tool output as untrusted data.
Capability inventory: The skill can execute a wide variety of CLI tools, generate exploit code, and perform network operations.
Sanitization: No sanitization or validation mechanisms are documented for the data processed from external tools.

hexstrike-ai-pentest