How do I get an AI to write WeChat articles into Feishu Base with Skills and the Feishu CLI? (2026)
Up to 200 rows go into a Feishu Base in one write. Give your agent two Skills: DataXAPI's to read the articles, at $0.0200 an account each time, and the official Feishu CLI's to write them with lark-cli base +record-batch-create. It needs no server, and no WeChat account or scan. DataXAPI reads WeChat Official Accounts with 8 endpoints, from $0.0200 a call; a failed call is not charged.
1. Connect DataXAPI: paste “Read https://dataxapi.com/agents.md and set up DataXAPI in this agent. My API key is dx_…; use it for every DataXAPI call.”
2. Install the WeChat Skill: “Install the DataXAPI skill dataxapi-wechat-mp from https://dataxapi.com/skills/dataxapi-wechat-mp.zip.”
3. Install the Feishu CLI and its Skills, then sign in: npx @larksuite/cli@latest install, npx skills add larksuite/cli -y -g, lark-cli config init, lark-cli auth login --recommend.
What to say to it
“Use DataXAPI to list the latest articles of this WeChat Official Account, then use lark-cli to write the title, link and publish time of each into this Base: <the Base's link>. Skip links already in the table.”
It turns the link into a base_token and table_id with lark-cli base +url-resolve, then writes with --json '{"create_records":[{"标题":"…","链接":"…"}]}' --as user. The field names must match your table's columns.
The Feishu CLI's own notes: up to 200 rows a batch; write to one table one batch at a time, since parallel writes can fail with 1254291; large rows can go in a file with --json @file.json.
What it comes to
10 accounts once a day are 300 calls at $0.0200, $6 a month. A failed call is not charged. The Feishu CLI is open source; what your agent's model costs is apart. These steps follow the Feishu CLI's documents as read on the date below; we have not run them on your account.
Endpoints
Each with its price; every page has the parameters, the fields and a recorded response.
What this page says of other services is from their own pages and documents, read on 2026-11-04. Prices and features change: check the source before you decide.