Automate Xiaohongshu note collection and publishing
2 ： RPA ，OpenClaw skill ， xiaohongshu-mcp 。 、 、 ， ， 。 Task: The author implemented a content pipeline using Yingdao RPA to scrape Xiaohongshu notes into Feishu Bitable, OpenClaw with a custom skill to generate compliant copy, and xiaohongshu-mcp for automated publishing. Human involvement is limited to initial setup, keyword configuration, and triggering the final publish command.; tools: OpenClaw. Human role: The original post does not say how a person stayed involved.; limitations: Requires 2 hours for deployment, specific RPA permissions, and Feishu API credentials. Token costs for OpenClaw are noted as high for full automation, hence the RPA hybrid approach. Platform compliance risks (banned words) are mitigated by a custom skill but not guaranteed..
Tools
Published on Agent Work Atlas:
Overview
2 ： RPA ，OpenClaw skill ， xiaohongshu-mcp 。 、 、 ， ， 。
Workflow
The author implemented a content pipeline using Yingdao RPA to scrape Xiaohongshu notes into Feishu Bitable, OpenClaw with a custom skill to generate compliant copy, and xiaohongshu-mcp for automated publishing. Human involvement is limited to initial setup, keyword configuration, and triggering the final publish command.
Reported result
The source claims the workflow saves 3 hours daily and enables 'one-sentence' publishing, but provides no verified metrics, error rates, or long-term stability data. Actual business impact remains unmeasured beyond the author's anecdotal success.
Human role
The original post does not say how a person stayed involved.
Limitations
Requires 2 hours for deployment, specific RPA permissions, and Feishu API credentials. Token costs for OpenClaw are noted as high for full automation, hence the RPA hybrid approach. Platform compliance risks (banned words) are mitigated by a custom skill but not guaranteed.
Try it
Act as an automation engineer. I need to replicate a Xiaohongshu content pipeline. 1. Configure Yingdao RPA to scrape [replaceable: target keywords] from Xiaohongshu and save structured data (title, body, stats) to [replaceable: Feishu Bitable URL]. 2. Set up OpenClaw with a custom skill that analyzes this data to generate compliant, engaging copy (title, body, tags) avoiding banned words. 3. Integrate xiaohongshu-mcp to allow publishing via a single command. First, perform a dry run on 5 sample notes to verify data integrity and copy quality. Do not publish anything until I manually approve the generated content. If scraping fails or API limits are hit, stop and report the error. Ensure all credentials are stored securely and not hardcoded.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.