What problem does it solve?
It turns natural-language requests about Xiaohongshu actions (login, publish, search, interact, and composite operations) into the correct, audited CLI workflows without the user manually figuring out which script/command to run.
Core Features & Use Cases
- Intent-based routing: Automatically maps user intents to sub-skills such as
xhs-auth, xhs-publish, xhs-explore, xhs-interact, and xhs-content-ops.
- Safety guardrails: Enforces that all Xiaohongshu operations must be executed only via
python scripts/cli.py <subcommand>, with explicit login checking (check-login) before actions.
- Structured outputs & user confirmation: Requires JSON-formatted CLI output and mandates user confirmation for publishing and commenting.
Use case example: You ask to “search notes about a keyword and then like a specific post after confirming details,” and the skill routes to xhs-explore and xhs-interact while ensuring login is checked first.
Quick Start
Ask your coding agent to “Search Xiaohongshu feeds for keyword ‘小红书运营’ and return the top results in JSON.”