IA_Core 🤖

Autonomous AI agent layer - Installs into any project and manages it intelligently using GPT-4o-mini (free tier) and OpenCore.

🚀 One-Line Installation

curl -fsSL https://raw.githubusercontent.com/YOUR_USERNAME/IA_core/main/install.sh | bash

That's it! IA_Core will:

Deeply analyze your project - Detect frameworks, languages, dependencies
Request necessary credentials - OpenAI API key (recommended), MCP configuration
Auto-configure - Generate project-specific workflows and settings
Deploy MCP servers - Memory, context, and tools for advanced capabilities
Start monitoring - Agent runs in background watching your project
Add CLI tools - iacore command available globally

📋 Requirements

Python 3.9+ (required)
OpenAI API Key (recommended for full functionality)
- Get one at: https://platform.openai.com/api-keys
- Free tier with GPT-4o-mini works perfectly
- Without API key: limited functionality with fallbacks

🔧 Interactive Configuration

During installation, you'll be prompted for:

OpenAI API Key (optional but recommended)
- Used for intelligent analysis
- GPT-4o-mini free tier is sufficient
- Configure later: iacore config
MCP Configuration (for backend/API projects)
- Enables memory, context, and tools servers
- Auto-detected for Django, FastAPI, Flask, Express
- Can skip if not needed

The installer analyzes your project and configures itself optimally!

🎯 Key Features

🤖 Fully Autonomous - Works independently after installation
🧠 GPT-4o-mini Powered - Free-tier LLM for intelligent analysis
👻 Invisible Execution - Commands via OpenCore (silent)
📊 Project-Aware - Understands 20+ project types automatically
🔄 Real-time Monitoring - Watches file changes intelligently
⚙️ Smart Configuration - Auto-configures based on project analysis
🔌 MCP Protocol - Memory, context, tools servers for advanced features
🎨 Rich CLI - Beautiful terminal UI with status and controls
🌐 REST API - HTTP API on port 8788 for integrations
💾 Persistent Learning - Remembers and improves over time
🔒 Secure - Credentials stored safely, dangerous commands blocked

📖 Usage

Once installed, simply work on your project normally. IA_Core observes and assists transparently.

CLI Commands

# View status and project analysis
iacore status

# See agent logs in real-time
iacore logs

# Pause/resume the agent
iacore pause
iacore resume

# Configure settings (API keys, workflows)
iacore config

# Re-analyze project after major changes
iacore analyze

# Check version
iacore version

# Uninstall completely
iacore uninstall

🔍 What Gets Analyzed

IA_Core performs deep project analysis during installation:

Frameworks: React, Vue, Angular, Next.js, Django, FastAPI, Flask, Express
Languages: Python, JavaScript, TypeScript, Go, Rust, Java
Build Tools: npm, pip, cargo, go modules, gradle
Dependencies: package.json, requirements.txt, go.mod, Cargo.toml
Structure: Components, modules, configuration
Recommended Agents: Based on detected technologies

Example analysis output:

📊 Project Analysis:
  • Type: Python
  • Frameworks: fastapi, sqlalchemy
  • Languages: python
  • Recommended Agents: backend-developer, api-engineer
  • MCP: ✅ Configured (memory + context + tools)

🏗️ Architecture

┌──────────────────────────────────────────┐
│       Your Project (any type)            │
├──────────────────────────────────────────┤
│         IA_Core (invisible layer)        │
│  ┌────────────┐      ┌────────────────┐ │
│  │  Detector  │─────▶│ IntelligentAI  │ │
│  └────────────┘      └────────────────┘ │
│  ┌────────────┐      ┌────────────────┐ │
│  │ GPT-4o-mini│◀─────│AutonomousAgent │ │
│  └────────────┘      └────────────────┘ │
│         │                    │           │
│         ▼                    ▼           │
│  ┌────────────┐      ┌────────────────┐ │
│  │   Memory   │      │   OpenCore     │ │
│  │   (MCP)    │      │   Executor     │ │
│  └────────────┘      └────────────────┘ │
└──────────────────────────────────────────┘

Components:

ProjectDetector: Analyzes project type, frameworks, dependencies
IntelligentAnalyzer: LLM-powered deep understanding of codebase
AutonomousAgent: Main loop monitoring files and executing workflows
LLMClient: GPT-4o-mini integration with rate limiting and caching
OpenCoreExecutor: Silent command execution in sandboxed environment
MCP Servers: Memory, context, and tools for advanced agent capabilities
FastAPI Server: REST API for external integrations
Typer CLI: Rich command-line interface for user control

🔌 MCP (Model Context Protocol)

IA_Core includes three MCP servers that enhance agent capabilities:

Memory Server

Persistent memory across sessions:

# Stores facts, decisions, learnings
memory/store_fact
memory/retrieve_fact
memory/store_decision
memory/store_learning
memory/get_learnings

Context Server

Project understanding and code navigation:

# Search, read, analyze code
context/search_files
context/read_file
context/find_definition
context/get_structure
context/project_summary

Tools Server

Safe command execution:

# Execute commands, modify files
tools/execute_command
tools/write_file
tools/search_code
tools/analyze_dependencies

MCP servers are automatically configured during installation for projects that need them (detected backend/API frameworks).

📊 Project Support Matrix

Type	Auto-Detection	Workflows	MCP
React	✅	Component analysis, build optimization	Optional
Vue	✅	Component analysis, state management	Optional
Angular	✅	Module analysis, dependency check	Optional
Next.js	✅	SSR optimization, route analysis	Optional
Python	✅	Code quality, type checking, tests	✅
Django	✅	API analysis, migrations, admin	✅
FastAPI	✅	Endpoint analysis, validation	✅
Flask	✅	Route analysis, blueprints	✅
Node.js	✅	Dependency audits, package updates	Optional
Express	✅	API endpoints, middleware analysis	✅
Go	✅	Module analysis, formatting	Optional
Rust	✅	Cargo check, clippy suggestions	Optional

🛠️ Advanced Configuration

Manual Configuration

Edit .iacore/config.yml in your project:

version: 1

project:
  type: python
  root: /path/to/project
  frameworks: [fastapi, sqlalchemy]

agent:
  enabled: true
  auto_analyze: true
  auto_execute: false  # Set true for full autonomy
  watch_mode: true

llm:
  provider: openai
  model: gpt-4o-mini
  config_file: ~/.iacore/llm_config.yml

mcp:
  enabled: true
  config_path: .iacore/mcp/config.yml
  auto_start: true

api:
  host: 127.0.0.1
  port: 8788
  auto_start: true

workflows:
  on_file_change:
    - detect_impact
    - analyze_context
  on_git_commit:
    - analyze_commit
    - suggest_improvements
  on_package_change:
    - update_dependencies
    - verify_security

Environment Variables

# OpenAI API Key
export OPENAI_API_KEY='sk-...'

# Custom project root for MCP servers
export PROJECT_ROOT='/path/to/project'

# Custom IA_Core home directory
export IACORE_HOME='~/.iacore'

📡 REST API

IA_Core exposes a REST API on http://127.0.0.1:8788:

# Check health
curl http://127.0.0.1:8788/health

# Get agent status
curl http://127.0.0.1:8788/status

# Analyze project
curl -X POST http://127.0.0.1:8788/analyze \
  -H "Content-Type: application/json" \
  -d '{"path": "."}'

# Pause/resume agent
curl -X POST http://127.0.0.1:8788/agent/pause
curl -X POST http://127.0.0.1:8788/agent/resume

See full API documentation: API_REFERENCE.md

🔒 Security

IA_Core is designed with security in mind:

✅ Sandboxed execution - Commands run in controlled environment
✅ Dangerous command blocking - rm -rf, sudo, shutdown blocked
✅ Credential encryption - API keys stored securely with proper permissions
✅ Rate limiting - Prevents API abuse (10 req/min on free tier)
✅ Audit logging - All actions logged for review
✅ No external data - Project data stays local

🤝 Contributing

Contributions welcome! See CONTRIBUTING.md for guidelines.

📄 License

MIT License - See LICENSE 682F for details.

🔗 Links

Documentation: docs/
Agent Spec: AGENT_SPEC.md
Deployment Guide: DEPLOYMENT.md
Issues: GitHub Issues
Discussions: GitHub Discussions

💬 Support

Need help?

Check the documentation
Review AGENT_SPEC.md for technical details
Search existing GitHub Issues
Create a new issue with details

Built with ❤️ for autonomous AI development

Name		Name	Last commit message	Last commit date
Latest commit History 7 Commits
bin		bin
iacore		iacore
skills		skills
.claude-plugin		.claude-plugin
.gitignore		.gitignore
AGENT_SPEC.md		AGENT_SPEC.md
CONTRIBUTING.md		CONTRIBUTING.md
DEPLOYMENT.md		DEPLOYMENT.md
EXECUTE_PUSH.md		EXECUTE_PUSH.md
GITHUB_SETUP.md		GITHUB_SETUP.md
INSTALLATION_FEATURES.md		INSTALLATION_FEATURES.md
LICENSE		LICENSE
README.md		README.md
SKILL.md		SKILL.md
create_and_push.sh		create_and_push.sh
fix_install_sh.sh		fix_install_sh.sh
fix_remote_and_push.sh		fix_remote_and_push.sh
init_and_push.sh		init_and_push.sh
install.sh		install.sh
manual_push.sh		manual_push.sh
mcp.json		mcp.json
push_to_github.sh		push_to_github.sh
requirements.txt		requirements.txt
setup.py		setup.py
setup_github.sh		setup_github.sh
test_install.sh		test_install.sh
verify.sh		verify.sh

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

IA_Core 🤖

🚀 One-Line Installation

📋 Requirements

🔧 Interactive Configuration

🎯 Key Features

📖 Usage

CLI Commands

🔍 What Gets Analyzed

🏗️ Architecture

🔌 MCP (Model Context Protocol)

Memory Server

Context Server

Tools Server

📊 Project Support Matrix

🛠️ Advanced Configuration

Manual Configuration

Environment Variables

📡 REST API

🔒 Security

🤝 Contributing

📄 License

🔗 Links

💬 Support

About

Uh oh!

Releases

Packages

Uh oh!

Languages

License

jeturing/IA_Core

Folders and files

Latest commit

History

Repository files navigation

IA_Core 🤖

🚀 One-Line Installation

📋 Requirements

🔧 Interactive Configuration

🎯 Key Features

📖 Usage

CLI Commands

🔍 What Gets Analyzed

🏗️ Architecture

🔌 MCP (Model Context Protocol)

Memory Server

Context Server

Tools Server

📊 Project Support Matrix

🛠️ Advanced Configuration

Manual Configuration

Environment Variables

📡 REST API

🔒 Security

🤝 Contributing

📄 License

🔗 Links

💬 Support

About

Resources

License

Contributing

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Languages

Packages