squidy 2.1.2

Smart Setup for AI Agent Projects - Governance and Audit

🦑 Squidy v2.1.2

Smart Setup for AI Agent Projects
Governance, Audit, and Automatic Documentation for Claude, GPT-4, Cursor, and more

Installation • Usage • Features • Documentation • Contributing

---

Star History

---

🎯 The Problem

You use Claude, ChatGPT, or Cursor to code, but:

• 🤯 AI forgets everything in the next conversation (context loss)
• 📝 You rewrite the same requirements every week
• 🎨 The agent gets "creative" and changes your architecture without warning
• 📂 Your project becomes a mess because no one documents anything
• ⏱️ You spend 30 min setting up prompts before you start coding

Squidy solves this in 2 minutes.

---

✨ What is Squidy?

Squidy is a premium CLI that automatically creates a governance structure for AI Agent projects.

🚀 Instead of typing 20 messages explaining your project every time:

1. 🎤 Chat with AI about your project (5-6 smart questions)
2. 🧠 Let the AI understand your stack, rules, and conventions
3. 📋 Get 10 documentation files ready to use
4. 🔒 Create a "contract" between you and the AI: rules, prohibitions, DoD

Result: Your AI never loses context or goes out of scope.

---

🎬 Demo

$ squidy init

🦑 AI Agent Setup

How it works:
  1. Tell me about your project (one sentence is enough)
  2. I'll ask up to 6 contextual questions
  3. Automatically generate 10 documentation files

🤖 Agent: Hi! Tell me about the project you want to configure.
         Example: "REST API for delivery with Node and PostgreSQL"
   You: REST API for delivery with Node and PostgreSQL

🤖 Agent: Great! Which frontend framework will you use?
   You: React with TypeScript

🤖 Agent: Perfect! Will you need authentication?
   You: Yes, JWT

... (2-3 more questions) ...

✅ Configuration generated successfully!

🦑 10 files generated at /home/user/delivery-api

Next steps:
  1. Tell your agent: "Access /home/user/delivery-api/readme-agent.md and follow the ritual"
  2. Review /home/user/delivery-api/doc/constitution.md
  3. Add tasks to /home/user/delivery-api/doc/kanban.md

---

📦 Installation

Via pip (recommended)

pip install squidy
squidy --version

Via pipx (isolated)

pipx install squidy
squidy --version

Development

git clone https://github.com/seomarc/squidyrun.git
cd squidyrun
python -m venv venv && source venv/bin/activate  # Linux/Mac
# or: python -m venv venv && venv\Scripts\activate  # Windows
pip install -e ".[dev]"
squidy --version

Requirements: Python 3.9+

---

🎮 Usage

Setup with AI (Recommended)

# Interactive setup with AI interview
squidy init

# Specify path
squidy init ./my-project

# Simulate without creating files (dry-run)
squidy init --dry-run

# Manual setup (without AI)
squidy init --manual

# Choose language (pt-BR or en-US)
squidy init --lang en-US

🌍 Supported Languages

Squidy v2.1+ supports multiple languages! All documentation and interface are generated in the selected language:

Language	Code	Status
🇧🇷 Portuguese (Brazil)	pt-BR	✅ Complete
🇺🇸 English (US)	en-US	✅ Complete

How to use:

# Select language via flag
squidy init --lang en-US

# Or let Squidy prompt you to choose
squidy init

# 🌍 Select your language:
# [1] 🇧🇷 Português (Brasil)
# [2] 🇺🇸 English (US)

📖 Complete Internationalization Guide

What is translated:

• ✅ Complete CLI interface
• ✅ All 10 documentation templates
• ✅ AI interview prompts
• ✅ Audit and status messages

Project Audit

# Audit current directory
squidy audit

# Audit specific project
squidy audit ./my-project

# JSON output
squidy audit -f json

# Apply automatic fixes
squidy audit --fix

Quick Status

# Show project status
squidy status

# Complete diagnosis
squidy doctor

---

🏗️ What Does Squidy Create?

Squidy generates a complete governance structure:

my-project/
├── readme-agent.md          # 🤖 Complete guide for the AI agent
├── .squidy/
│   └── manifest.json        # 📋 Project manifest
├── doc/
│   ├── AGENT.md             # 🎯 Quick reference for the agent
│   ├── constitution.md      # ⚖️  Principles, prohibitions, DoD
│   ├── oracle.md            # 🧙 Architecture decisions (ADRs)
│   ├── policies.md          # 📋 Stack, conventions, policies
│   ├── kanban.md            # 📊 Task management (Epics → Tasks → Subtasks)
│   ├── emergency.md         # 🚨 Critical blocker registry
│   ├── diary-index.md       # 📑 History index
│   └── session-context.md   # 💾 Current state cache
└── diary/
    └── 2026-02.md           # 📅 Automatic decision log

📋 Kanban Structure

## 🔥 EPICS
### EPIC-001: Authentication System
**Priority:** P0 | **Complexity:** M
**Tasks:** TASK-001, TASK-002

## 📋 BACKLOG
### TASK-001: Setup JWT [EPIC-001]
**Complexity:** S | **Priority:** P0
**Subtasks:**
- [ ] SUB-001: Install library (XS - 30min)
- [ ] SUB-002: Configure middleware (S - 1h)

## 🏗️ IN PROGRESS (WIP: 1/3)
- [ ] TASK-001: Setup JWT

## ✅ COMPLETED
- [x] TASK-000: Initial setup

---

🎨 Features

✨ v2.1 - New Features

• 🌍 Multi-language - Full support for Portuguese and English (pt-BR, en-US)
• 📋 Bilingual Templates - Documentation generated in selected language
• 🤖 Multi-language AI - Interview and configuration in Portuguese or English

✨ v2.0 - New Features

• 🎨 Premium UI/UX - Modern interface with Rich, gradients, and animations
• 🤖 Smart Interview - 5 structured phases with contextual follow-ups
• 📊 Complete Audit - Checks structure, kanban, freshness, and consistency
• 🔌 Clean Architecture - Ports & Adapters, extensible and testable
• 📋 Templates v2.0 - More complete and actionable documentation
• 🧪 Tests - pytest suite, 10/10 passing

🤖 Supported AI Providers

Provider	Models	Cost
OpenAI	GPT-4o-mini	Paid
Anthropic	Claude 3 Haiku/Sonnet	Paid

🔍 Audit

Squidy can audit existing projects:

• ✅ StructureChecker - Checks required files
• ✅ KanbanChecker - Analyzes WIP limit, blocked tasks
• ✅ FreshnessChecker - Identifies outdated files
• ✅ ConsistencyChecker - Checks consistency between files

---

🛠️ Tech Stack

• Python 3.9+ - Main language
• Typer - CLI framework
• Rich - UI components and formatting
• Pydantic v2 - Data validation
• Jinja2 - Templates
• OpenAI / Anthropic - AI providers

---

📚 Documentation

• 📖 Complete Documentation
• 🌍 Internationalization Guide - Multi-language (pt-BR, en-US)
• 🚀 Quick Start Guide
• 🏗️ Architecture
• 🤝 Contributing
• 🌐 Contributing with Translations

---

🤝 Contributing

Contributions are welcome! Read our Contributing Guide.

Development

# Clone
git clone https://github.com/seomarc/squidyrun.git
cd squidyrun

# Setup
python -m venv venv && source venv/bin/activate
pip install -e ".[dev]"

# Tests
pytest

# Lint
black squidy/
isort squidy/
mypy squidy/

# Commit
pre-commit run --all-files

---

🔗 Links

🌐 Website • 📦 PyPI • 💻 GitHub • ▶️ YouTube • 💼 LinkedIn • ☕ Buy Me a Coffee

👤 Developer

• Marcos Tadeu - Personal Website
• SearchOps - searchops.io

---

💖 Support the Project

If Squidy helped you, consider:

• ⭐ Give a star on GitHub
• 🐦 Share on Twitter
• 💼 Use it at your company
• 🤝 Contribute with code
• ☕ Buy Me a Coffee

---

📄 License

MIT License - see LICENSE for details.

---

Made with 🦑 by Marcos Tadeu • 💻 Project GitHub

🌐 squidy.run • 📦 PyPI • 💻 GitHub • ▶️ YouTube • 💼 LinkedIn • 👤 Developer • 🏢 SearchOps • ✉️ Contact