Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
53 changes: 53 additions & 0 deletions agents/TilmanGriesel__chipper/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,53 @@
# Chipper

**Chipper** is a modular, hackable, privacy-first RAG (Retrieval-Augmented Generation) agent platform for tinkerers, educators, and builders who want powerful local LLM workflows without sending data to the cloud.

---

## Run

```bash
npx @open-gitagent/gitagent run -r https://github.com/TilmanGriesel/chipper
```

---

## What It Can Do

| Skill | Description |
|---|---|
| **RAG Query** | Answer questions grounded in your documents via ElasticSearch vector retrieval |
| **Document Embed** | Chunk and vectorize documents into the knowledge base |
| **Web Scrape** | Extract and index content from web pages |
| **Audio Transcribe** | Convert audio files to text and index the transcript |
| **Ollama API Proxy** | Proxy any Ollama client request with server-side RAG and system prompt injection |

---

## Key Features

- **Local & Cloud Models** — Run models via [Ollama](https://ollama.com/) locally or connect to [Hugging Face](https://huggingface.co/) API
- **ElasticSearch Integration** — Scalable, vectorized document storage and retrieval
- **Fully Dockerized** — One `docker compose up` and it's running
- **Ollama API Compatible** — Works as a drop-in proxy for Enchanted, Open WebUI, Ollamac, and the Ollama CLI
- **Offline Web UI** — Built with vanilla JS + TailwindCSS, no internet required
- **Edge TTS** — Listen to responses using WebAssembly-based client-side TTS
- **Distributed Processing** — Chain multiple Chipper instances together

## Why Chipper?

Chipper was born from a real need: to help explore a book's characters and ideas using local LLMs, keeping private creative work entirely off the cloud. That origin gives it a clear soul — **privacy-first, approachable, and built to be forked**.

**Live demo:** https://demo.chipper.tilmangriesel.com/
**Project site:** https://chipper.tilmangriesel.com/

## Setup

```bash
git clone https://github.com/TilmanGriesel/chipper.git
cd chipper
python setup.py # interactive setup
docker compose up -d # launch all services
```

See the [full setup guide](https://chipper.tilmangriesel.com/get-started.html) for detailed instructions.
14 changes: 14 additions & 0 deletions agents/TilmanGriesel__chipper/metadata.json
Original file line number Diff line number Diff line change
@@ -0,0 +1,14 @@
{
"name": "chipper",
"author": "TilmanGriesel",
"description": "Hackable, privacy-first RAG agent platform — local LLMs via Ollama, ElasticSearch vector store, doc chunking, web scraping, audio transcription, and an Ollama API proxy with server-side RAG injection.",
"repository": "https://github.com/computer-agent/chipper",
"version": "1.0.0",
"category": "developer-tools",
"tags": ["rag", "ollama", "local-llm", "haystack", "elasticsearch", "privacy", "docker", "tinkerer", "knowledge-base", "ollama-proxy"],
"license": "MIT",
"model": "ollama:llama3",
"adapters": ["system-prompt", "openai"],
"icon": false,
"banner": false
}