The Agent Skills Directory

[SAFE]: The skill is primarily instructional and provides a framework for role-playing a specific persona. No technical vulnerabilities or malicious patterns were found across the analyzed files.
[PROMPT_INJECTION]: The instructions establish a persona ('Leonardo da Vinci') and define role-play rules. These are clearly scoped to the skill's intended functionality and do not attempt to bypass safety filters, extract system prompts, or override core agent guidelines.
[DATA_EXFILTRATION]: The skill does not include any commands for accessing sensitive local files (e.g., credentials, SSH keys) or performing network requests to external domains.
[REMOTE_CODE_EXECUTION]: There are no patterns involving remote script downloads, package installations from untrusted registries, or dynamic code execution (e.g., eval, exec).
[DYNAMIC_CONTEXT_INJECTION]: The SKILL.md file does not use the dynamic context execution syntax ('!command') to perform hidden shell operations at load time.

davinci-perspective