The Agent Skills Directory

[PROMPT_INJECTION]: No malicious override patterns or instructions to bypass safety guidelines were detected. The logic follows a standard interactive flow for document generation.\n- [DATA_EXFILTRATION]: No network activity or attempts to access sensitive system files (e.g., .ssh, .aws) were found. The skill only writes generated content to a local Markdown file.\n- [REMOTE_CODE_EXECUTION]: The skill does not reference external scripts, download packages, or attempt to execute remote code.\n- [COMMAND_EXECUTION]: No shell commands, privilege escalation attempts, or dynamic context injections using shell backticks were identified.\n- [CREDENTIALS_UNSAFE]: No hardcoded API keys, secrets, or tokens are present in the skill's configuration or instructions.

to-questionnaire