GitHub Repository

A ohmyzsh plugin that automatically debugs failed terminal commands using a local LLM via Ollama. Local-only, privacy-first, and focused on one thing: **fix what just broke**. No cloud. No API keys. No background daemons.

9 starsShell

Sleuther, a ohmyzsh plugin for local LLM debugging with Ollama

Name: Sleuther, a ohmyzsh plugin for local LLM debugging with Ollama
Availability: InStock
Author: chocks

by chocks·Mar 23, 2026·2 points·0 comments

Visit Project View on HN

AI Analysis

●●SolidNiche GemBig Brain

Local LLM debugging with sanitization fallback before cloud share — clever privacy approach.

Strengths

•Sanitization strips usernames, paths, and secrets before suggesting cloud fallback.
•Suggested commands blocked from pipes, redirection, and destructive operations by default.
•Zero API keys or background daemons — truly local-first workflow.

Weaknesses

•Zsh hooks don't expose stdout/stderr, limiting context for auto-mode debugging.
•Requires Ollama setup and model management, adding friction for casual users.

Post Description

Recently, I noticed my debugging workflow had become: copy error into ChatGPT/Claude -> go back and repeat (while being cautious not copying over sensitive stuff like paths, names etc.)

So, I built Sleuther, a Oh My Zsh plugin that asks a local LLM via Ollama to explain and debug errors directly in the terminal. Right now I am using Qwen2.5-Coder.

At a previous job, I used warp.dev extensively and was missing its agent mode. Sleuther is inspired by warp and zsh-copilot.

No copy paste, no cloud, no context switching.

Also, has a small trick: if the local LLM cannot help after a couple attempts, it suggests a sanitized version of the error, stripping usernames, paths, and hostnames before sharing with ChatGPT or Claude. The sanitization is rather rudimentary right now but something I want to improve.

Curious if anyone else is experimenting with local LLMs for dev workflows?

https://github.com/chocks/sleuther