xhs-job-helper

Automated Xiaohongshu job posting collection uses time-based keyword rotation to search feeds, filters out non-recruitment content, and skips already-seen note IDs to ensure structured deduplication into a persistent markdown dataset.

The two-stage daily push prevents duplicate job digests by preparing pending items, verifying delivery status, and then committing pushed IDs using a two-phase commit with seen-ids and pending-push-ids tracking.

You can query collected Xiaohongshu recruitment data using natural language to filter by company, keyword, or date range, reading from data.md and archived files to output formatted summaries or JSON.

Scheduling Xiaohongshu job scraping for morning, afternoon, and evening is handled through scheduled cron collection, executing script-driven JSON pipelines to systematically gather and store recruitment posts.

You need Python uv installed to run the Xiaohongshu job digest pipeline, as it manages the script-driven JSON pipelines required for data collection, markdown parsing, and delivery status verification.

Missing or duplicated Xiaohongshu job posts in the daily digest occur when delivery status verification fails, preventing the two-phase commit from correctly updating seen-pushed-ids and pending-push-ids.