rednote(xiaohongshu,xhs)

Comprehensive tool for interacting with rednote (xiaohongshu,小红书) platform. This skill enables users to search for posts by keyword, extract content from spe...

This skill appears to do what it claims: automated browser interactions with Xiaohongshu. Before installing or running it: (1) review the included Python scripts yourself or run them in an isolated environment (VM or container); (2) be aware the scripts will save your login cookies to rednote_cookies.json — treat that file like a credential (store it securely or use a throwaway account); (3) automated actions (likes, comments, publishes) will run as your account and may violate platform terms — proceed with caution; (4) ensure you trust the source before installing Playwright and running arbitrary scripts; (5) if you want stronger safety, modify scripts to run headless-only or add confirmation prompts before performing write actions.

Type: OpenClaw Skill Name: rednote-skills Version: 1.0.0 The skill bundle is classified as suspicious due to several vulnerabilities that could be exploited via prompt injection against the OpenClaw agent. Specifically, `scripts/publish_note.py` directly uses user-provided `--image-urls` for file uploads, creating a risk of arbitrary local file exfiltration or RCE if an attacker can control these arguments. Similarly, `scripts/comment_note.py` directly uses `comment_text`, posing a potential XSS risk if the platform is vulnerable and the agent's input is not sanitized. The `rednote_cookies.json` file, which stores authentication tokens, is also a sensitive asset that, while not directly exfiltrated by the skill, could be targeted by an attacker manipulating the agent.