yoetz
Fast, CLI-first LLM council + bundler + multimodal gateway for coding agents.
Note: This project is under active development. APIs may change.
Why yoetz?
Most LLM CLI tools focus on a single provider or a single workflow. yoetz is different:
- Multi-model council — get consensus from multiple LLMs in one command, not sequential copy-paste between tabs
- Multimodal native — text, images, and video as first-class inputs across providers
- Agent-first design — structured JSON output, local budget tracking for ask/council/review, and agent skill integration out of the box
- Zero lock-in — one config, any provider (OpenRouter, OpenAI, Gemini, LiteLLM), switch with a flag
- Bundle-aware — package your codebase with gitignore-awareness for maximum LLM context
Table of Contents
Features
- Bundle: Package code files with gitignore-awareness for LLM context
- Ask: Query LLMs with text, images, or video
- Council: Multi-model consensus with configurable voting
- Review: AI-powered code review for diffs and files
- Generate: Create images (OpenAI) and videos (Sora, Veo)
- Browser: Fallback to web UIs via recipes
Installation
Homebrew (macOS / Linux)
brew install avivsinai/tap/yoetz
Scoop (Windows)
scoop bucket add avivsinai https://github.com/avivsinai/scoop-bucket
scoop install yoetz
Pre-built Binaries
Download the latest release from GitHub Releases.
# macOS (Apple Silicon)
curl -LO https://github.com/avivsinai/yoetz/releases/latest/download/yoetz-aarch64-apple-darwin.tar.gz
tar xzf yoetz-aarch64-apple-darwin.tar.gz
sudo mv yoetz /usr/local/bin/
# macOS (Intel)
curl -LO https://github.com/avivsinai/yoetz/releases/latest/download/yoetz-x86_64-apple-darwin.tar.gz
tar xzf yoetz-x86_64-apple-darwin.tar.gz
sudo mv yoetz /usr/local/bin/
# Linux (x86_64)
curl -LO https://github.com/avivsinai/yoetz/releases/latest/download/yoetz-x86_64-unknown-linux-gnu.tar.gz
tar xzf yoetz-x86_64-unknown-linux-gnu.tar.gz
sudo mv yoetz /usr/local/bin/
# Linux (ARM64)
curl -LO https://github.com/avivsinai/yoetz/releases/latest/download/yoetz-aarch64-unknown-linux-gnu.tar.gz
tar xzf yoetz-aarch64-unknown-linux-gnu.tar.gz
sudo mv yoetz /usr/local/bin/
From Source
cargo install --git https://github.com/avivsinai/yoetz
Build Locally
git clone https://github.com/avivsinai/yoetz.git
cd yoetz
cargo build --release
Agent Skill (Claude Code / Codex CLI)
# Via skills marketplace
/plugin marketplace add avivsinai/skills-marketplace
/plugin install yoetz@avivsinai-marketplace
# Via skills.sh
npx skills add avivsinai/yoetz
# Via skild.sh
npx skild install @avivsinai/yoetz
Quick Start
Configuration
Create ~/.yoetz/config.toml:
[defaults]
provider = "openrouter"
model = "anthropic/claude-sonnet-4-5-20250929"
[providers.openrouter]
api_key_env = "OPENROUTER_API_KEY"
[providers.openai]
api_key_env = "OPENAI_API_KEY"
[providers.gemini]
api_key_env = "GEMINI_API_KEY"
Basic Usage
# Bundle files for LLM context
yoetz bundle --prompt "Review this code" --files "src/**/*.rs"
# Ask a question
yoetz ask --prompt "Explain this function" --files "src/main.rs"
# Ask with structured JSON output (OpenAI-compatible)
yoetz ask --prompt "Return JSON only" --provider openai --model gpt-5.2 --response-format json
# Ask with an image (vision)
yoetz ask --prompt "Describe this diagram" --image diagram.png --provider gemini --model gemini-3-flash-preview
# Override MIME type for signed/extensionless URLs
yoetz ask --prompt "Describe this" --image https://example.com/signed --image-mime image/png
# Ask about a video
yoetz ask --prompt "Summarize this" --video meeting.mp4 --provider gemini --model gemini-3-flash-preview
# Override video MIME type for signed/extensionless URLs
yoetz ask --prompt "Summarize this" --video https://example.com/signed --video-mime video/mp4
> Note: Gemini can return empty content if `--max-output-tokens` is too low because tokens are consumed by thoughts. If you see warnings or empty output, increase the limit.