DuckTapeKiller

DuckTapeKiller / horme

Public

Minimalist, privacy-first AI assistant for Obsidian. Powered by local LLMs by default, requiring no API keys and ensuring zero data leaks. Includes an optimised Vault Brain (RAG) with character-offset indexing and automated privacy firewalls to secure data if optional cloud models are enabled.

ai obsidian-plugin ollama privacy rag-chatbot
12
0
100% credibility
Found May 09, 2026 at 12 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Horme is an Obsidian plugin that adds a privacy-focused AI chat assistant to interact with, search, edit, and enhance your personal notes using local or optional cloud models.

How It Works

1
📖 Discover Horme

You find this handy AI sidekick for your Obsidian notes app while browsing helpful add-ons.

2
⬇️ Add it to Obsidian

Download a few simple files and drop them into your notes app's add-on folder, then turn it on in settings.

3
🤖 Prepare your local thinker

Grab two small brain models on your computer so the AI can work privately without sending anything online.

4
⚙️ Unlock note memory

Flip the switch to let it remember and search all your notes automatically in the background.

5
💬 Chat about your notes

Open the chat window and ask anything – it instantly finds and pulls in the perfect bits from your notes to help.

6
✏️ Polish your writing

Right-click any text to proofread, rewrite in a new style, summarize, or check facts with one tap.

🎉 Smarter notes forever

Your entire knowledge collection now thinks with you, suggesting tags, summaries, and insights – all safe and private.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 12 to 12 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is horme?

Horme is a TypeScript Obsidian plugin that turns your vault into a privacy-first AI assistant, powered by local LLMs like Ollama—no API keys or data leaks required. It indexes your notes via a lean Vault Brain RAG system for semantic search, letting you query your knowledge base directly in a chat panel or via right-click tools for proofreading, rewriting, and fact-checking. Upload PDFs or DOCX for context, generate frontmatter summaries, and use built-in skills for Wikipedia lookups or date calculations, all while keeping everything on-device by default.

Why is it gaining traction?

Unlike bloated AI plugins relying on cloud APIs, Horme prioritizes local models with a privacy firewall that blocks vault data from cloud providers unless you opt-in, appealing to devs wary of leaks. Its right-click menu delivers instant editing (rewrite in formal tone, translate, beautify Markdown) and semantic tagging from thousands of vault tags, blending seamlessly into Obsidian workflows. Like a minimalist GitHub readme or UI Lovelace setup, it strips fluff for fast, token-aware chats with exportable history.

Who should use this?

Obsidian power users building personal knowledge bases who want local AI for querying notes without sending data to OpenAI or Claude. Researchers or writers proofreading Spanish grammar from custom manuals, or PKM enthusiasts linking related notes semantically. Avoid if you're not running Ollama—cloud fallback exists but limits RAG.

Verdict

Try it if you're deep in Obsidian and value privacy; the README rivals polished minimalist GitHub profiles with clear setup for nomic-embed-text indexing. At 12 stars and 1.0% credibility, it's early-stage raw but feature-complete—watch for maintenance as hormel-like niche tools (hormesis in PKM?) mature.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.