Swarmauri builds open source developer tools for modular AI applications, agent workflows, governed software delivery, and reusable automation. The ecosystem centers on the Swarmauri SDK, a Python toolkit for composing models, tools, agents, parsers, embeddings, vector stores, and application workflows through consistent interfaces.

Install the main Python package:

pip install swarmauri

Install the full SDK surface:

pip install "swarmauri[full]"

Use the documentation for API references, examples, and package guidance:

• Swarmauri SDK documentation
• Swarmauri SDK repository
• Swarmauri on PyPI

Swarmauri projects are designed for developers who need AI systems that can be assembled, tested, extended, and operated as real software.

Use these entry points when choosing where to start:

• Build with the SDK: swarmauri-sdk
• Explore notebooks: swarmauri-notebooks
• Generate projects and files: peagen
• Try a retrieval chatbot: rag_assistant
• Prototype with an LLM playground: swarmauri-playground
• Build interface components: swarmakit
• Use CRUD router patterns: crouton
• Manage multi-package Python repositories: soliloquy

Most Swarmauri projects are Python-first and work well with pip, uv, and virtual environments. Start with the repository README for the package you are using, then move to the hosted docs for API details and examples.

Common paths:

python -m pip install swarmauri
python -m pip install "swarmauri[full]"
python -m pip install peagen

For contributors:

git clone https://github.com/swarmauri/swarmauri-sdk.git
cd swarmauri-sdk
uv sync --all-extras --dev

• Modular packages: install only the interfaces, base classes, integrations, or tools you need.
• Consistent contracts: build against stable component interfaces instead of one-off provider code.
• Extensible workflows: compose agents, tools, parsers, vector stores, and model integrations in predictable ways.
• CLI-friendly operations: keep generation, validation, release, and automation flows scriptable.
• Traceable delivery: use governed artifacts, tests, evidence, and reusable workflows where reliability matters.

Open an issue or pull request in the relevant repository with a focused description, reproduction steps when applicable, and the package or workflow surface involved. For SDK usage, include the package name, Python version, installation command, and the smallest code sample that shows the behavior.