The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides specific command-line instructions for the agent to execute, such as using the Hydra tool to perform brute-force attacks against target login endpoints (Phase 4, SKILL.md).- [DYNAMIC_EXECUTION]: The skill includes an embedded Python script in Phase 6 that utilizes the requests library to perform multiple automated network calls to a user-specified target for the purpose of session token collection and analysis.- [INDIRECT_PROMPT_INJECTION]: The skill processes external application data (HTTP responses, headers, and tokens), which introduces a potential attack surface if the target application is controlled by an adversary.
Ingestion points: The agent is instructed to capture and analyze HTTP responses and session tokens from target URLs (SKILL.md).
Boundary markers: The instructions do not define clear delimiters or include warnings to help the agent distinguish between its instructions and potentially malicious data returned by the application.
Capability inventory: The skill allows the agent to perform network requests (Python) and execute system commands (Hydra), providing a vector for automated exploitation (SKILL.md).
Sanitization: The skill lacks specific instructions for sanitizing or validating external input before processing or reporting results to the user.

broken-authentication