Set up Little Heta for the first time.
heta initWhat it does:
- Creates
~/.heta/heta.yaml. - Configures one LLM provider: Qwen, ChatGPT, Gemini, or custom.
- Optionally configures MinerU for PDF and Office parsing.
- Enables vector indexing and insert planning by default.
- Installs the Little Heta skill into Codex and Claude Code.
Prepare before running:
- Your LLM provider API key.
- For custom providers, prepare chat and embedding API settings. Model names
with a provider prefix such as
openai/gpt-5.4-nanoare treated as LiteLLM-native and do not require a base URL. Bare model names require an OpenAI-compatible/v1base URL. - Optional custom multimodal settings if you want image parsing.
- Optional MinerU API key from https://mineru.net/apiManage/docs.