API Reference PipelineEx v#0.1.1

AI pipeline orchestration library for Elixir.

Manages checkpoint creation and restoration for pipeline execution.

Unified AST parsing utilities for different programming languages.

Specialized analyzer for Elixir projects.

Specialized analyzer for Go projects.

Specialized analyzer for JavaScript/TypeScript projects.

Specialized analyzer for Python projects.

Specialized analyzer for Rust projects.

Codebase discovery and context analysis system.

Core discovery engine for codebase analysis.

Core query engine for intelligent codebase analysis.

Enhanced conditional execution engine supporting complex boolean expressions and mathematical operations.

Function library for advanced condition expressions.

Configuration management system for the pipeline.

Context management for nested pipeline execution.

Data transformation engine for manipulating structured data between pipeline steps.

Debug logging utilities for the pipeline.

Debugging tools and utilities for nested pipeline execution.

Enhanced configuration management for Claude Agent SDK integration.

Enhanced error formatting and context tracking for nested pipeline execution.

Main pipeline execution engine.

The genetic blueprint structure for pipeline organisms. Defines the DNA schema used by the Genesis Pipeline and evolution system.

Manages environmental adaptation capabilities

Determines execution patterns and runtime behavior

Controls performance optimization characteristics

Controls the fundamental architecture of the pipeline

Pipeline generator that transforms DNA into executable pipeline configurations. This is the core engine that brings pipeline organisms to life.

Performance metrics collection and analysis for nested pipeline execution.

Performance monitoring and metrics collection for pipeline execution.

Registry for pipeline monitoring processes.

Streaming-specific performance metrics collection and analysis.

Pipeline wrapper for ClaudeAgentSDK.OptionBuilder with additional pipeline-specific functionality.

Automated benchmark runner for comparing sync vs async streaming performance.

Builds prompts from template configurations. Includes file caching for improved performance.

Behavior for AI service providers (Claude, Gemini, etc.)

Live Claude provider built on top of ClaudeAgentSDK.

Extended Claude provider with configurable timeout support. This module works around the hardcoded 30-second timeout in ClaudeAgentSDK.

Enhanced Claude provider with full Claude Agent SDK integration.

Live Gemini provider backed by the gemini_ex client.

Manages step results throughout pipeline execution.

Recursion protection and circular dependency detection for nested pipelines.

Resource monitoring and management for nested pipeline execution.

Unified safety management for nested pipeline execution.

Schema for basic analysis responses from Gemini.

Common JSON Schema definitions for pipeline step outputs.

Session management for persistent Claude conversations.

Manages pipeline variables and state interpolation.

Step execution coordination module.

Claude step executor - handles all Claude (Muscle) operations.

Claude Batch step executor - handles claude_batch step type with parallel processing capabilities.

Claude Extract step executor - handles claude_extract step type with content extraction and processing.

Enhanced Claude step type for robust execution with advanced error recovery patterns.

Claude Session step executor - handles claude_session step type with session management integration.

Claude Smart step executor - handles claude_smart step type with OptionBuilder preset integration.

Codebase query step executor for intelligent code analysis and querying.

Data transformation step executor for manipulating structured data between pipeline steps.

File operations step executor for comprehensive file management.

Gemini step executor - handles all Gemini (Brain) operations using InstructorLite.

Executes Gemini (Brain) steps using InstructorLite for structured output and function calling.

Advanced loop step executor - handles for_loop and while_loop operations.

Executes another pipeline as a step within the current pipeline.

Executes multiple Claude tasks in parallel.

Set variable step executor - handles variable assignment and state management.

Simple test step that echoes its value for testing purposes. Only available in test environment.

Analyzes streaming performance and provides optimization recommendations.

Result streaming capabilities for pipeline execution.

Behaviour for mocking AI providers (Gemini) in tests.

Mock support for testing async streaming functionality.

Behaviour for mocking Claude SDK in tests.

Behaviour for mocking file system operations in tests.

Common test utilities for pipeline testing.

Behaviour for mocking Logger operations in tests.

Mock implementations for testing pipeline components without external dependencies.

Mock implementation of Claude provider for testing.

Mock implementation of Claude Agent SDK for testing.

Mock implementation of file system operations for testing.

Mock implementation of Gemini AI provider for testing.

Mock implementation of Gemini provider for testing.

Mock implementation of Logger for testing.

Mock implementation of Session Manager for testing claude_session steps.

Manages test mode configuration for switching between mock and live providers.

Adapter for using Pipeline tools with InstructorLite.

Tool to get the WAN (external) IP address using Ubuntu 24.04 CLI tools.

Base behavior for pipeline tools that can be called by LLMs.

Registry for managing and executing pipeline tools.

Execution tracing and monitoring for nested pipeline execution.

File utility functions for pipeline file operations.

JSON Schema-based validation for pipeline step outputs.

Run performance benchmarks comparing sync vs async streaming.

Generates a new pipeline using the Genesis Pipeline.

Generates a new pipeline using the Genesis Pipeline in LIVE mode.

Runs a Pipeline workflow from a YAML configuration file.

Runs a Pipeline workflow from a YAML configuration file in LIVE mode.

Demonstrates the Pipeline with both mock and live modes.

Runs Pipeline tests in live mode with real API calls.