LLM Woodpecker

LLM Woodpecker is an AI model debugging tool that supports both Chat mode and Raw JSON mode.

Features

• Chat Mode: Interactive chat interface for testing LLM responses
• Raw JSON Mode: Direct JSON request/response testing with structured view
• Multiple Model Support: Configure and switch between different LLM providers
• Streaming Support: Real-time streaming responses
• Modern UI: Clean, responsive interface with collapsible content

Installation

pip install llm-woodpecker

Usage

Web UI

llm-woodpecker

Then open http://127.0.0.1:7660 in your browser.

Configuration

Initialize config file:

llm-woodpecker --init-config

List configured models:

llm-woodpecker --list-models

Set active model:

llm-woodpecker --set-model qwen

Chat Mode (Terminal)

llm-woodpecker --chat

Or with a message:

llm-woodpecker --chat --messages '[{"role":"user","content":"Hello!"}]'

Configuration File

The default config file is located at ~/.llm_woodpecker/config.yaml.

License

MIT License