Frequently Asked Questions

ECC (The Agent Harness Performance Optimization System) is a comprehensive system for AI development that provides skills, instincts, memory optimization, security scanning, and research-first development. It works with Claude Code, Codex, Opencode, Cursor and other AI agent harnesses to enhance productivity and code quality.

Recommended method (Plugin):

/plugin marketplace add https://github.com/affaan-m/ECC
/plugin install ecc@ecc
                            

Manual installation:

git clone https://github.com/affaan-m/ECC.git
cd ECC
npm install
./install.sh --profile full
                            

For detailed instructions, see the Installation Guide.

ECC supports multiple platforms:

• Claude Code: Native plugin support with full feature parity
• Codex: Full support for both macOS app and CLI
• Cursor IDE: Complete IDE integration with hooks and rules
• OpenCode: Plugin support with 20+ hook events
• GitHub Copilot: Instruction and prompt layer integration
• Zed: Project-level support with adapters
• Gemini CLI: Experimental project-local support

• 67 Specialized Agents: For planning, architecture, code review, security, testing
• 271 Skills: Pre-built workflows for TDD, continuous learning, verification
• AgentShield Security: 102 static analysis rules and OWASP audits
• Memory Optimization: Context persistence across sessions
• Continuous Learning: Auto-extract patterns from sessions
• Cross-Platform Support: Works across 12+ language ecosystems
• Dashboard GUI: Desktop dashboard with visual component exploration

ECC includes multiple security features:

• AgentShield Integration: Built-in security scanning with 102 static analysis rules
• Secret Detection: Automatic detection of API keys and secrets
• Sandboxing: Restricted execution environment for risky operations
• OWASP Audits: Comprehensive security checklist following OWASP Top 10
• Input Sanitization: Automatic sanitization of user inputs
• Security Review: Dedicated security review agent

Use the /security-scan command for quick security checks.

ECC supports 12+ language ecosystems:

• JavaScript/TypeScript: React, Next.js, Node.js patterns
• Python: Django, Flask, pytest patterns
• Go: idiomatic patterns, testing with Go testing
• Java: Spring Boot, Quarkus, Maven/Gradle
• Swift: iOS development with modern patterns
• PHP: Laravel patterns and security
• Perl: Modern Perl 5.36+ patterns
• Rust: idiomatic patterns and build resolution
• Kotlin: Android/KMP development
• HarmonyOS/ArkTS: Mobile app development
• C++: Core guidelines and testing patterns

Contributions are welcome! You can contribute by:

• Creating new agents or skills
• Improving existing workflows
• Adding support for new platforms
• Contributing to documentation
• Reporting bugs and issues
• Becoming a sponsor

See CONTRIBUTING.md for guidelines. Check out our Community Discussions for collaboration.

ECC is licensed under the MIT License. This means you can:

• Use the software for any purpose
• Modify the source code
• Distribute the software
• Use it in commercial projects

The only requirement is including the original license and copyright notice in your distributions. See the full LICENSE file for details.

There are several ways to get support:

• Discord Community: Join our Discord server for real-time discussions
• GitHub Discussions: Ask questions and share ideas in GitHub Discussions
• Documentation: Check our comprehensive guides and documentation
• GitHub Issues: Report bugs and request features via GitHub Issues
• Sponsors: For premium support, consider sponsoring the project

ECC stands out with:

• Cross-Harness Design: Works across multiple AI coding platforms
• Performance Optimization: Built from real-world multi-harness workflows
• Complete System: Not just configs - includes skills, memory, security
• Research-First: Continuous learning and pattern extraction
• Security Focus: Built-in security scanning and protection
• Extensibility: Plugin architecture for customization
• Community-Driven: Active development with 230+ contributors