Naming Convention: extensions/¶

Overview¶

This document describes the extensions/ directory in the aipartnerupflow project.

Directory Purpose¶

extensions/ - Framework Extensions¶

Purpose: Framework-provided, production-ready optional extensions

Characteristics: - ✅ Production-grade implementations - ✅ Full test coverage - ✅ Maintained by framework maintainers - ✅ Can have heavy dependencies - ✅ Installed via extras: [crewai], [stdio] - ✅ Direct use in production - ✅ Registered through ExtensionRegistry system

Examples: - extensions/crewai/ - CrewAI executor (LLM tasks) - extensions/stdio/ - Stdio executor (local command execution) - extensions/core/ - Core executors (aggregate_results_executor) - extensions/hooks/ - Hook implementations - extensions/storage/ - Storage implementations - extensions/tools/ - Tool implementations

Location: src/aipartnerupflow/extensions/

Learning Resources¶

For examples and learning templates, see the test cases: - tests/integration/ - Integration tests demonstrating real-world usage patterns - tests/extensions/ - Extension-specific test examples

Test cases serve as comprehensive examples showing how to: - Use executors in real scenarios - Create custom task implementations - Build task trees with dependencies - Aggregate results from multiple tasks

Extension System¶

All extensions in extensions/ must: 1. Implement ExecutableTask interface (or extend BaseTask) 2. Use @extension_register() decorator for auto-registration 3. Be registered in ExtensionRegistry by unique ID 4. Follow the unified extension system architecture

See extension-registry-design.md for details.

Summary¶

• extensions/: Framework-provided, production-ready executors (CrewAI, Stdio, etc.)
• Test cases: Serve as learning resources and examples (see tests/integration/ and tests/extensions/)

Use extensions/ for production-ready, reusable components. For learning and examples, refer to the test cases.