The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands and network utilities such as curl to probe web endpoints for redirection behavior. It also uses security tools like nuclei for automated vulnerability scanning.
[EXTERNAL_DOWNLOADS]: Installs the openredirex package from the official Python Package Index (PyPI) to provide specialized redirection testing capabilities.
[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from external files which could contain malicious payloads designed to subvert the agent's logic. * Ingestion points: Reads from recon/$TARGET/urls.txt and recon/$TARGET/redirect-candidates.txt which typically contain data gathered from external web crawling. * Boundary markers: Implements a mandatory confirmation gate requiring the user to explicitly verify target URLs and commands before execution. * Capability inventory: Includes network access (curl), package management (pip3), and file system interaction within the recon/ directory. * Sanitization: No explicit sanitization or filtering of the input URL data is performed before processing.

hunt-open-redirect