The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a Python script to process markdown files into HTML. The shell commands described in the documentation (python3 scripts/build.py) are standard for executing local scripts and do not involve any risky operations or privilege escalation.- [EXTERNAL_DOWNLOADS]: The skill does not perform any network requests or downloads. It relies exclusively on the Python standard library and local template assets.- [DATA_EXFILTRATION]: There is no evidence of data exfiltration. The script processes local files or stdin and writes the output to a local path specified by the user. The browser-side JavaScript includes an 'Export to Markdown' feature which generates a local download via a Blob URL, remaining entirely on the client side.- [PROMPT_INJECTION]: The instructions do not contain patterns typical of prompt injection or attempts to bypass AI safety filters. It provides clear, functional guidance for formatting interview questions.- [INDIRECT_PROMPT_INJECTION]: While the skill processes untrusted user data (interview questions), the processing logic involves basic string parsing and HTML escaping. The script uses html.escape to mitigate potential Cross-Site Scripting (XSS) when rendering the HTML output. The impact is limited to the generated static HTML file.

meet-record-html