The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/chrome_launcher.py uses subprocess.Popen to launch a Google Chrome instance with remote debugging enabled. This allows the skill to control the browser via the Chrome DevTools Protocol (CDP).
[DYNAMIC_EXECUTION]: The skill makes extensive use of the Runtime.evaluate CDP command in scripts/cdp_publish.py and scripts/feed_explorer.py to execute JavaScript within the browser context. This is the primary method used for DOM manipulation and data extraction from the Xiaohongshu website.
[EXTERNAL_DOWNLOADS]: The scripts/image_downloader.py module uses the requests library to download media files (images and videos) from external URLs provided by the user. These files are saved to temporary local directories for subsequent upload to Xiaohongshu.
[DATA_EXFILTRATION]: While the skill accesses and manages local browser profiles and configurations in config/accounts.json, this data is used strictly for local automation and profile isolation. No evidence of unauthorized data transmission to external servers was found.
[PROMPT_INJECTION]: The instructions in SKILL.md are standard for tool usage and do not contain any patterns intended to bypass AI safety guidelines or override system behavior.
[SAFE]: The skill's operations are transparent and align with its stated goal of providing automation for Xiaohongshu. Although it possesses powerful browser control capabilities, these are implemented through standard CDP protocols and are used for legitimate automation tasks.

RedBookSkills