Dropshipping, but make it conversational
An AI skill that lets you chat your way from 1688 supplier to TikTok shop owner.

What it does
1688-shopkeeper is a Python CLI and OpenClaw skill that bridges China’s largest wholesale platform (1688) with four downstream Chinese e-commerce stores: Douyin, Pinduoduo, Xiaohongshu, and Taobao. You search for products, inspect details, and publish inventory—all through natural language prompts or structured JSON commands.
The interesting bit
The project treats e-commerce operations as an agent skill rather than a dashboard. The CLI outputs a consistent JSON envelope with a markdown field for human reading and a data field for programmatic use, which suggests it was built to be consumed by an LLM orchestrator from day one. The “chat to dropship” UX is the product, not a wrapper.
Key highlights
- Natural language interface: “find summer dresses for Douyin” → search → “publish to my shop”
- Supports full pipeline on four platforms: product search, detail lookup, and one-click publishing
- Includes trend analysis, opportunity heatmaps, and daily shop reports
- CLI returns structured JSON with markdown explanations for dual human/agent consumption
- Configuration requires 1688 AI app and an access key; no standalone API signup path
Caveats
- Requires downloading a separate 1688 AI mobile app to obtain an access key; no purely web or API-based onboarding
- “Limited-time free” pricing for core features; actual costs are opaque and tied to 1688 platform rules
- All support channels route through the 1688 AI app, not GitHub issues
Verdict
Worth a look if you’re building LLM-native commerce tools for the Chinese market or exploring agent-driven SMB workflows. Skip it if you need transparent pricing, REST API stability guarantees, or operate outside 1688’s ecosystem.