llcat 0.2.0

/usr/bin/cat for the LLM era

/usr/bin/cat for LLMs

llcat is an LLM program with very little ambition in life.

That's what makes it useful.

You can handle this!

llcat solves all your problems. Yes, all of them!

You can also:

• Pipe things into it or prompt it on the command line.
• Conversation history as an optional file.
• Tool Calling with the OpenAI spec. There's a file and a program example in this repository.
• Servers:
  • OPENAI_API_BASE and LLM_BASE_URL are supported along with -s for one off.
  • Authentication tokens are passed with -k. You can do $(< somefile) or whatever obfuscation you want, that's on you.
• Models are listed with the -m option and specified with it as well.

Want it?

• pipx install llcat
• uv tool run llcat

Dependency? Just the requests library. That's it.

Examples

List the models on openrouter

llcat -s https://openrouter.ai/api -m

Go ahead, you can run that, unauthenticated, no problem.

$ llcat -s https://openrouter.ai/api \
        -m meta-llama/llama-3.2-3b-instruct:free \
        -c /tmp/convo.txt \
        -k $(cat openrouter.key) \
        "What is the capital of France?"

$ llcat -s https://openrouter.ai/api \
        -m meta-llama/llama-3.2-3b-instruct:free \
        -c /tmp/convo.txt \
        -k $(cat openrouter.key) \
        "And what about Canada?"

Let's continue it locally

$ llcat -s http://192.168.1.21:8080 \
        -c /tmp/convo.txt \
        "And what about Japan?"

Pure sorcery.

The Tool Call To Rule Them All

This uses the sophisticated 21 line example_tool_program.py included in this repository.

It also uses DA`/50's pretty little streaming markdown renderer, streamdown.

Kablam! Alright a16z where's my $50 million?

The enterprise applications are limitless...

Full documentation

usage: llcat [-h] [-c CONVERSATION] [-m [MODEL]] [-k KEY] [-s SERVER]
                [-tf TOOL_FILE] [-tp TOOL_PROGRAM]
                [prompt ...]

positional arguments:
  prompt                Your prompt

options:
  -h, --help            show this help message and exit
  -c, --conversation CONVERSATION
                        Conversation history file
  -m, --model [MODEL]   Model to use (or list models if no value)
  -k, --key KEY         API key for authorization
  -s, --server SERVER   Server URL (e.g., http://::1:8080)
  -tf, --tool_file TOOL_FILE
                        JSON file with tool definitions
  -tp, --tool_program TOOL_PROGRAM
                        Program to execute tool calls