Common utilities, data structures, and pipeline definitions used across all ARDoCo modules.

Core pipeline definition and orchestration for ARDoCo.

Model provider stage for loading and processing software architecture models.

Text extraction stage for extracting and processing textual information from documentation.

Recommendation generator stage for creating traceability link recommendations.

Text preprocessing stage for natural language processing of documentation using Stanford CoreNLP.

JSON schema definition and implementation for the TextProvider interface.

Connection generator stage for creating connections between textual and model elements.

A Java library that uses Large Language Models to automatically identify and extract architectural entities (like components and interfaces) from Software Architecture Documentation. It supports multiple LLM providers (OpenAI, Ollama) with ...

Base classes and utilities for testing ARDoCo modules. Includes benchmark data from https://github.com/ardoco/benchmark.

Core inconsistency detection stage for identifying missing or deviating elements between documentation and models.