The Agent Skills Directory

[SAFE]: The skill is designed for interactive idea validation and does not exhibit malicious patterns. It does not perform remote downloads, exfiltrate sensitive data, or attempt to escalate privileges.
[COMMAND_EXECUTION]: The skill utilizes an internal orchestration script (.agents/scripts/orchestrator_state.js) to manage project state. This execution is limited to local scripts provided by the vendor and is used for project lifecycle management.
[PROMPT_INJECTION]: The skill includes instructions to guide agent behavior through specific questioning techniques. These instructions are intended to refine user input and do not attempt to bypass core safety guardrails or extract system information.

validate-idea