Yoetz Skill

Fast, agent-friendly LLM council tool for multi-model consensus, code review, and bundling.

When to Use

Explicit triggers only:

• •"yoetz ask" / "yoetz council" / "yoetz review"
• •"yoetz bundle" / "yoetz generate" / "yoetz browser"
• •"use yoetz to..."

NOT triggered by:

• •"second opinion" / "ask another model" (could be amq-cli)
• •"council" alone / "review" alone (other skills may apply)

Installation (auto-bootstrap)

Before running any yoetz command, ensure the CLI is installed. If command -v yoetz fails, install via one of the following:

Prefer Homebrew when available — pre-built binaries, fastest install.

Agent Contract

• •Always use --format json for parsing
• •Set YOETZ_AGENT=1 environment variable
• •Parse JSON results and present summary to user
• •For large bundles, run yoetz bundle first to inspect size

Quick Reference

See commands reference for full details.

Council (Multi-Model Consensus)

Get opinions from multiple LLMs in parallel. --models is required.

yoetz council \
  -p "Should we use async traits or callbacks for this API?" \
  -f src/lib.rs -f src/api/*.rs \
  --models openai/gpt-5.2,gemini/gemini-pro-3,openrouter/xai/grok-4.1 \
  --format json

Example council sets:

• •Cross-provider: openai/gpt-5.2,gemini/gemini-pro-3,openrouter/xai/grok-4.1
• •Via OpenRouter only: openrouter/openai/gpt-5.2,openrouter/anthropic/claude-sonnet-4,openrouter/google/gemini-pro-3

Ask (Single Model)

Quick question with file context:

yoetz ask \
  -p "What's the bug in this error handling?" \
  -f src/error.rs \
  --provider openai --model gpt-5.2 \
  --format json

For Anthropic/XAI models, use OpenRouter (no extra config needed):

yoetz ask -p "Review this" -f src/*.rs \
  --provider openrouter --model anthropic/claude-sonnet-4 \
  --format json

Review

Staged changes

yoetz review diff --staged --format json

Specific file

yoetz review file --path src/main.rs --format json

With custom model

yoetz review diff --staged --provider openai --model gpt-5.2 --format json

Bundle (for manual paste or browser mode)

Bundle creates a session with files at ~/.yoetz/sessions/<id>/bundle.md.

# Get bundle path from JSON output
yoetz bundle -p "Explain this" -f src/**/*.rs --format json
# Output includes: {"artifacts":{"bundle_md":"/Users/.../.yoetz/sessions/.../bundle.md",...},...}

# Extract bundle_md path directly
BUNDLE=$(yoetz bundle -p "Review" -f src/*.rs --format json | jq -r .artifacts.bundle_md)
cat "$BUNDLE"

For browser workflows, pass the bundle.md path:

BUNDLE=$(yoetz bundle -p "Review" -f src/*.rs --format json | jq -r .artifacts.bundle_md)
yoetz browser recipe --recipe recipes/chatgpt.yaml --bundle "$BUNDLE"

Browser Fallback (Experimental)

For web-only models like ChatGPT Pro that lack API access. Uses Oracle-style cookie extraction from real Chrome to bypass Cloudflare challenges.

Prerequisites

# Node >=22 required. agent-browser is auto-resolved via npx if not in PATH.
# Install sweet-cookie for cookie extraction
npm install -g @steipete/sweet-cookie

Profile location

Default profile dir: ~/.config/yoetz/browser-profile/

Override per machine:

export YOETZ_BROWSER_PROFILE=/path/to/profile

First-time setup

Step 1: Log into ChatGPT in real Chrome

1. •Open Chrome (the real browser, not Playwright)
2. •Navigate to https://chatgpt.com/
3. •Log in with your account
4. •Close Chrome completely

Step 2: Sync cookies to agent-browser

yoetz browser sync-cookies

This extracts your authenticated cookies from Chrome and saves them for agent-browser. State file is stored at ~/.config/yoetz/browser-profile/state.json (or your overridden profile path).

Step 3: Verify authentication

yoetz browser check

Re-sync when sessions expire

If you see Cloudflare challenges or login prompts, re-sync:

# Log into ChatGPT in real Chrome, close Chrome, then:
yoetz browser sync-cookies
yoetz browser check

Use ChatGPT Pro via recipe

# Create bundle and get bundle.md path
BUNDLE=$(yoetz bundle -p "Review this code" -f src/*.rs --format json | jq -r .artifacts.bundle_md)

# Send to ChatGPT
yoetz browser recipe --recipe recipes/chatgpt.yaml --bundle "$BUNDLE"

Combined workflow: API + Browser

# Get fast API results first
yoetz council -p "Review" -f src/*.rs \
  --models openai/gpt-5.2,gemini/gemini-pro-3 --format json > api.json

# Then get ChatGPT Pro opinion
BUNDLE=$(yoetz bundle -p "Review" -f src/*.rs --format json | jq -r .artifacts.bundle_md)
yoetz browser recipe --recipe recipes/chatgpt.yaml --bundle "$BUNDLE"

How it works

The browser module uses stealth techniques to avoid Cloudflare detection:

• •Extracts real cookies from Chrome's encrypted cookie store
• •Injects them into Playwright via --state
• •Uses realistic User-Agent headers
• •Disables automation detection flags (--disable-blink-features=AutomationControlled)

Provider Configuration

Built-in providers (work with just env var):

• •openai - OPENAI_API_KEY
• •gemini - GEMINI_API_KEY
• •openrouter - OPENROUTER_API_KEY

Via OpenRouter (recommended for Anthropic/XAI - no extra config):

• •openrouter/anthropic/claude-sonnet-4
• •openrouter/xai/grok-4.1

Model format: provider/model

• •openai/gpt-5.2
• •gemini/gemini-pro-3
• •openrouter/anthropic/claude-sonnet-4 (nested for OpenRouter)

Cost Control

# Estimate before running
yoetz pricing estimate --model gpt-5.2 --input-tokens 12000 --output-tokens 800

# Set limits
yoetz ask -p "Review" --max-cost-usd 1.00 --daily-budget-usd 5.00 --format json

Tips

• •Use --debug to capture raw responses during troubleshooting
• •Gemini may return empty content if --max-output-tokens is too low
• •Session artifacts stored in ~/.yoetz/sessions/<id>/
• •For image inputs: yoetz ask -p "Describe" --image photo.png --format json
• •ChatGPT recipe placeholder may vary by locale; check snapshot output if fill fails