The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes multiple curl command templates in SKILL.md and a dedicated Python script scripts/agent.py for performing network-based security assessments against user-specified targets.
[SAFE]: No malicious patterns were identified. The network operations performed by the script and commands are consistent with the skill's stated purpose of web application security testing. There is no evidence of credential harvesting, unauthorized file access, or data exfiltration.
[SAFE]: The Python script scripts/agent.py uses the well-known requests library for HTTP communications and does not employ any dangerous dynamic execution patterns such as eval() or exec() on untrusted data.

performing-http-parameter-pollution-attack