goal is to pass a git repo to the LLM so it can consume and we can quary it.
we will use a cli tool called llm
and an auxiliry tool called files-to-prompt from same auther
install llm
pip install llm
llm install llm-openrouter
#auth, if you have $OPENAI_API_KEY in your enviroment, and want to use gpt you can use llm without setting keys
llm keys set openrouter #paste key when asked
llm models default claude-3.5-sonnet #or gpt-4o for openai
we will get llm.exe
tool we can run in any terminal
llm "five names for a gaint pinguin baby"
install files-to-prompt
pip install files-to-prompt
note the pipe |
symbol, it passes the output of one command to another.
git clone github.com/some/repo
files-to-prompt.exe .\repo\ | llm --system "in the code what port is used, and what address libcurl uses"
after inital push we should use the -c
to --continue
the conversation with prior knowledge
- if you have
$OPENAI_API_KEY
in your enviroment you can usellm
without any mods. - to set gpt-4o as default,
llm models default gpt-4o
- You can view the current model by running this:
llm models default