lingyao-ai
v0.2.2
Published
Lingyao setup CLI for Codex and Claude Code.
Readme
lingyao-ai
A customizable setup CLI that behaves like npx leishen-ai, with your own configuration.
Features
- Configure Codex and/or Claude Code in one interactive flow.
- Prompt for gateway URL first, then API key, with a config default.
- Optional auto-install checks for
@openai/codexand@anthropic-ai/claude-code. - API key validation supports
sk-andcr_formats. - Writes and backs up:
~/.claude/config.json~/.codex/config.toml~/.codex/auth.json~/.codex/.env- shell RC block (
.zshrc,.bashrc, etc.)
- On Windows, optionally patch VSCode ChatGPT extension model order.
1) Customize your config
Edit project.config.json:
brand.*: product name and shell marker.apiKey.*: API key format and env variable name.endpoints.*: your gateway domain and paths.cli.*: min Node version, npm mirror, package names.claude.*/codex.*: provider defaults and target files.codex.windowsVscodePatch.*: Windows extension patch behavior.
2) Run locally
cd /Users/espresso/Project/lingyao-ai
node cli.jsUse a custom config file:
node cli.js --config /absolute/path/to/project.config.json3) Publish and use with npx
Package name is already set to lingyao-ai.
npm publish --access publicThen users can run:
npx lingyao-aiDuring setup, the CLI will ask for a gateway URL first. Press Enter to use
project.config.json's default endpoints.baseOrigin, or enter a different
http:// / https:// URL before providing the API key.
