TrustyAI Explainability Python is a Python bindings library that wraps TrustyAI's explainability capabilities, enabling Python developers to access explainable AI functionality through a native Python interface. The library is maintained at version 0.6.3 and serves as the Python entry point to TrustyAI's broader explainability ecosystem, which is part of the Kogito project. The repository is classified across multiple AI trustworthiness domains including explainable AI, model explanations, AI fairness, AI robustness, and specific explanation techniques such as counterfactuals, SHAP, and LIME, as well as capabilities for detecting data drift and analyzing black-box models.

The library provides multiple installation pathways to accommodate different user needs. Users can install directly from PyPi for standard usage, set up dependencies locally from the repository root for development work, or use Docker containerization for isolated environments with Jupyter server access on localhost:8888. Additionally, the project supports interactive exploration through Binder, allowing users to run example Jupyter notebooks directly in the browser without local installation. Comprehensive API documentation and usage examples are available through the ReadTheDocs page, which serves as the primary reference for developers integrating the library into their projects.

The project maintains active development with a median issue and pull request response latency of 32 hours, though some items have experienced longer resolution times with a mean latency of 2536.4 hours. The most frequently tracked issue categories are bug reports with three instances, followed by dependency updates and enhancement requests with two instances each. This activity pattern reflects ongoing maintenance and incremental feature development rather than rapid iteration cycles.

The contributor ecosystem shows concentrated activity around core maintainers, with ruivieira leading engagement across 13 tracked events, followed by christinaexyou with 3 events and AmberJBlue with 1 event. This distribution indicates a relatively small but dedicated core team managing the project. The repository is part of a broader TrustyAI ecosystem, with overlapping contributors also active in related projects including the main trustyai-explainability repository, the trustyai-service-operator, and opendatahub-io/notebooks, suggesting coordinated development across multiple complementary tools and integrations.

The library is positioned as a bridge between Python developers and TrustyAI's Java-based explainability engine, making advanced model interpretability techniques accessible to the Python data science community. Working examples are maintained in a separate examples repository, allowing users to learn through practical demonstrations. The project welcomes community contributions through a documented contribution process outlined in the CONTRIBUTING.md file, supporting the collaborative development model typical of open source explainability tools.