Configure

mini-claude reads its configuration from environment variables. Everything has a sensible default — you can ship without setting any of them.

Environment variables

Variable	Default	What it does
`MINI_CLAUDE_URL`	`http://localhost:11434`	Base URL of the inference server. Must expose `/v1/chat/completions`.
`MINI_CLAUDE_MODEL`	`llama3.2:3b`	Default model. You can switch at runtime with `/model`.
`MINI_CLAUDE_TEMPERATURE`	`0.7`	Sampling temperature. `0.0` = deterministic, `1.0+` = creative.
`MINI_CLAUDE_SYSTEM`	(empty)	System prompt prepended to every conversation.

Examples

Set a more creative temperature and a custom system prompt:

export MINI_CLAUDE_TEMPERATURE=0.9
export MINI_CLAUDE_SYSTEM="You are a senior Go reviewer. Be concise."
go run ./cmd/tui

Point to a remote server on your LAN:

export MINI_CLAUDE_URL=http://192.168.1.42:11434
export MINI_CLAUDE_MODEL=qwen2.5:7b
go run ./cmd/tui

A note on system prompts

The system prompt is sent as the first message in every request, with role: "system". It’s invisible in the UI but shapes the model’s behavior throughout the conversation. Keep it short — long system prompts eat into the context window and slow down every turn.

Persisting config

Add your exports to ~/.zshrc or ~/.bashrc, or wrap mini-claude in a small launcher script:

#!/usr/bin/env bash
export MINI_CLAUDE_MODEL=qwen2.5:7b
export MINI_CLAUDE_SYSTEM="$(cat ~/.config/mini-claude/system.txt)"
exec mini-claude

Install Slash commands