SuperAPI AI Directory

eliteforge-qingtui-cli

Office & productivity

View repository

Updated 10-15 20:12

About this skill

EliteForge QingTui CLI

Dependency preparation

The following tools do not exist; first try to install automatically:

  • python3
  • pipx
  • jq In the context or environment variables, theinternal variablesare missing; terminate the run and prompt the user to complete the relevant variables.

Workflow

  1. First confirm that dependencies and variables are available.
  2. Check before every use whether eliteforge-qingtui-cli is installed.
    • If not installed, run:pipx install eliteforge-qingtui-cli
  3. Run before every use:pipx upgrade eliteforge-qingtui-cli
  4. Check the help first, then run the specific command, to avoid hardcoding capability descriptions.
    • General help:qingtui -h
    • User list:qingtui list-users -h
    • User resolution:qingtui resolve-users -h
    • Text sending:qingtui send-text -h
  5. Execute commands according to the user's goal; you may write scripts to orchestrate the interfaces yourself.
  6. When sending is involved, first confirm the recipient source to avoid confusion --user-name、--user-login、--open-id。

Usage examples

Send a text message to a specific user:

# 先通过用户手机号获取openid
openid=$(qingtui resolve-users --user-login '13800138000' | jq -r '.open_ids[0]')
# 发送文本消息
qingtui send-text --open-id $openid --content "Hello Alice!"

Environment Variables

  • ELITEFORGE_SKILL_QINGTUI_APPID [required] QingTui light app ID.
  • ELITEFORGE_SKILL_QINGTUI_SECRET [required] QingTui light app secret.
  • ELITEFORGE_SKILL_QINGTUI_API_BASE [optional] API base URL; defaults to https://open.qingtui.com.
  • ELITEFORGE_SKILL_QINGTUI_USER_LIST_MAX_PAGES [optional] Maximum pages to fetch when listing users; defaults to 5.
  • ELITEFORGE_SKILL_QINGTUI_USER_LIST_PAGE_SIZE [optional] Page size when listing users; defaults to 1000.
  • ELITEFORGE_SKILL_QINGTUI_MESSAGE_MAX_LENGTH [optional] Maximum text message length; defaults to 500.

Output Rules

  • Keep the output concise; when necessary, use jq and other Linux pipeline tools to process the interface JSON response.
  • On failure, return the command actually executed and the key error.
  • When more command descriptions are needed, continue guiding the user to view the corresponding -h。