The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected. The skill's functionality is limited to conversational interaction and text generation.
[PROMPT_INJECTION]: The skill contains no instructions to bypass safety filters or ignore system guidelines. It follows a structured interview flow to assist the user.
[DATA_EXFILTRATION]: There are no network request capabilities or mechanisms to access and transmit sensitive local data like credentials or SSH keys.
[COMMAND_EXECUTION]: The skill does not invoke subprocesses or execute shell commands. It explicitly states it is a read-only conversational skill.
[EXTERNAL_DOWNLOADS]: No external dependencies, scripts, or remote code patterns are present in the skill definition.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests user input to generate prompt templates (files, error messages, URLs), it lacks dangerous capabilities (code execution or network access) that would allow an injection to escalate into a system compromise. The skill also includes explicit instructions to sanitize inputs for shell metacharacters and directory traversal.

Prompt Wizard