- Source
jarvis.shin your.bashrcor.bash_profileor zshrc:source /path/to/jarvis/jarvis.sh - Ensure you have the
llmPython package installed (see original README for details). You have the choice to use any llm package that supports, but by default I configured to use llm-grok -- where you get abundant free token to use if you sign up with XAI to sell some of your data: https://docs.x.ai/docs/data-sharing.
- Installing llm https://llm.datasette.io/en/stable/setup.html
- Installing llm-grok https://github.com/Hiepler/llm-grok
Just run @jarvis with your command, and follow the instruction.
❯ @jarvis set a cron to run on computer start: ls -ltra
👋 I have some suggestion to your request:
a|👍 Accept) Run: (crontab -l 2>/dev/null; echo "@reboot ls -ltra") | crontab -
d|🖐️ Deny) Run: set a cron to run on computer start: ls -ltra
s|📑 Save) Save command to history, so you can edit and run later.
Enter your choice (a/s/d): a
===================CMD EXEC STARTED==================
===================CMD EXEC ENDED: 0==================
jarvis on feature/simple [✘!?] took 9s
❯ crontab -l
@reboot ls -ltra
If @jarvis thinks the command is right, or llm fails, then it will execute the command directly!
If execution fails, @jarvis also analyze it and gives you suggestions.
❯ JARVIS_LLM_MODEL=gpt-4o-mini @jarvis check my disk size
👋 I have some suggestion to your request:
a|👍 Accept) Run: df -h
d|🖐️ Deny) Run: check my disk size
s|📑 Save) Save command to history, so you can edit and run later.
Enter your choice (a/s/d): d
===================CMD EXEC STARTED==================
bash: check: command not found
===================CMD EXEC ENDED: 127==================
[🤖 ℹ️ ] The command failed because it is not recognized. To check your disk size, you can use the following command instead:
``bash
df -h
``
This will display the disk usage in a human-readable format.
By default, @jarvis uses grok-3-mini-fast-latest as the llm model. Make sure the command llm -m grok-3-mini-fast-latest 'what is the model of you' mentions about Elon Musk.
You can set the llm model by setting the environment variable JARVIS_LLM_MODEL:
export JARVIS_LLM_MODEL="gpt-4o-mini"You can use local llm as well, make sure testing if it works.
JARVIS_DEBUG = 1 # show debug JARVIS_DEBUG = 2 # show trace