oops-cli 0.1.0

Fix your last failed shell command with an LLM. Type `oops`.

oops

Type oops after a command fails. It figures out the fix and runs it.

LLM-powered command-line repair. Zero dependencies. Works with a local model — no API key, fully offline.

---

$ git statsu
git: 'statsu' is not a git command. See 'git --help'.

$ oops
🤖 thinking...

  git status

Run this? [Y/n] y
On branch main
nothing to commit, working tree clean

You mistyped a command, it failed, you sigh — and instead of squinting at the error you just type oops.

Install

pip install oops-cli
oops --install      # adds a tiny shell hook to your .bashrc / .zshrc
source ~/.bashrc    # or just open a new terminal

That's it. After any failed command, type oops.

Pick a brain

oops auto-detects a provider:

Provider	Setup	Notes
Anthropic	export ANTHROPIC_API_KEY=...	Fast, smart, default if the key is set
Local (Ollama)	install Ollama, ollama pull llama3.2	No key, fully offline, free

If ANTHROPIC_API_KEY is set it uses that; otherwise it falls back to a local Ollama model.

How it works

1. The shell hook hands oops the command you just ran.
2. oops re-runs it to capture the exact error output.
3. It asks the model for a corrected command.
4. It shows you the fix and runs it only if you say yes.

Nothing executes without your confirmation.

Config

Env var	Default	What
ANTHROPIC_API_KEY	–	Use the Anthropic API
OOPS_MODEL	claude-haiku-4-5-...	Anthropic model
OOPS_OLLAMA_MODEL	llama3.2	Local model name
OOPS_OLLAMA_URL	http://localhost:11434	Ollama endpoint

Why

The error message almost always contains the answer. oops just reads it for you so you don't have to context-switch. It's the difference between a 200ms oops and a 30-second detour to your browser.

Contributing

Issues and PRs welcome. Good first ones: more providers (OpenAI, Gemini, Groq), a --yes auto-run flag, multi-suggestion mode, and a --explain mode that explains instead of fixing.

License

MIT