The Agent Skills Directory

[SAFE]: The skill establishes a high security baseline by explicitly forbidding the use of JavaScript, external stylesheets, web fonts, and remote images in the generated HTML files, preventing common web-based attack vectors.- [SAFE]: Instructions in SKILL.md mandate the escaping of all input text and strictly prohibit placing source content in executable contexts or event handlers, which effectively mitigates Cross-Site Scripting (XSS) risks when viewing derived documents.- [SAFE]: The workflow incorporates integrity checks using SHA-256 hashes and requires explicit user confirmation before replacing existing files, adhering to safe file management practices.- [SAFE]: The skill operates entirely on local files for the purpose of document visualization, with no evidence of network exfiltration, hardcoded credentials, or unauthorized command execution.

visualize