Skip to main content

Python Bindings

Repository: https://github.com/liquidos-ai/AutoAgents

The Python bindings exist to make it easy to explore new ideas quickly without giving up the Rust core that powers AutoAgents.

Python stays focused on agent definition, orchestration, experimentation, and callback glue. Rust owns the runtime, scheduling, streaming, concurrency, and core execution model.

Stable vs Experimental

The production path is:

  • Rust-owned executors (BasicAgent, ReActAgent)
  • Rust-owned memory (SlidingWindowMemory and future Rust-backed providers)
  • Python tools
  • Python hooks

Explicit extension paths live under autoagents.experimental:

  • ExperimentalAgentBuilder
  • CustomExecutor
  • Python-backed memory adapters

Those extension APIs are experimental in the Python bindings specifically because they rely on Python-native execution or Python-native memory behavior. That boundary exists to keep production runtime ownership in Rust rather than re-centering the architecture on Python callbacks.

The same categories of capability are production-grade in the Rust-native implementation. What is experimental here is the Python extension path, not the underlying AutoAgents model when implemented natively in Rust.

Why This Matters

  • Rapid iteration in Python is useful when you are prototyping agent flows, hooks, memory adapters, and pipeline layers.
  • Performance-sensitive and safety-critical behavior still comes from the Rust implementation underneath.
  • Moving a successful prototype to production Rust is easier because the Python layer is orchestration-oriented rather than a separate runtime.

Practical Outcome

You can prototype in Python with:

  • the same LLMProvider model
  • the same pipeline composition model
  • the same agent builder structure
  • the same runtime concepts used by the Rust crates

When you decide to harden something for production, the path to Rust is much smaller because the architecture is already aligned.

Installation from PyPI

Install the base package:

pip install autoagents-py

Install with a local backend or guardrails using extras:

pip install "autoagents-py[llamacpp]"           # llama.cpp CPU
pip install "autoagents-py[llamacpp-cuda]"      # llama.cpp CUDA
pip install "autoagents-py[llamacpp-metal]"     # llama.cpp Metal (macOS)
pip install "autoagents-py[llamacpp-vulkan]"    # llama.cpp Vulkan
pip install "autoagents-py[mistralrs]"          # mistral-rs CPU
pip install "autoagents-py[mistralrs-cuda]"     # mistral-rs CUDA
pip install "autoagents-py[mistralrs-metal]"    # mistral-rs Metal (macOS)
pip install "autoagents-py[guardrails]"         # Guardrails
pip install "autoagents-py[llamacpp-cuda,guardrails]"  # Multiple extras

Local Development

From the repository root:

uv venv --python=3.12
source .venv/bin/activate
uv pip install -U pip "maturin>=1.13.3,<2" pytest pytest-asyncio pytest-cov

make python-bindings-build

For local backend work:

make python-bindings-build-llamacpp-only
make python-bindings-build-mistralrs-only

Backend-specific distributables now live in dedicated package directories under bindings/python/ such as autoagents-llamacpp-cuda and autoagents-mistralrs-metal, so local builds and release builds use the same checked-in package metadata.

For full setup and usage details, see:

  • README.md
  • docs/src/getting-started/python-bindings.md

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

autoagents_guardrails_py-0.4.0.tar.gz (496.0 kB view details)

Uploaded Source

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

autoagents_guardrails_py-0.4.0-cp39-abi3-manylinux_2_28_x86_64.whl (8.9 MB view details)

Uploaded CPython 3.9+manylinux: glibc 2.28+ x86-64

autoagents_guardrails_py-0.4.0-cp39-abi3-manylinux_2_28_aarch64.whl (8.5 MB view details)

Uploaded CPython 3.9+manylinux: glibc 2.28+ ARM64

autoagents_guardrails_py-0.4.0-cp39-abi3-macosx_11_0_arm64.whl (8.1 MB view details)

Uploaded CPython 3.9+macOS 11.0+ ARM64

File details

Details for the file autoagents_guardrails_py-0.4.0.tar.gz.

File metadata

  • Download URL: autoagents_guardrails_py-0.4.0.tar.gz
  • Upload date:
  • Size: 496.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for autoagents_guardrails_py-0.4.0.tar.gz
Algorithm Hash digest
SHA256 66ed173d5a9f66e381f6da59a1b8c65d2507e922e3c49ea4eb3cb3738ad05708
MD5 fb5384cb0f779df69243ffb93b724bd8
BLAKE2b-256 2d5506c451ddb023400477f1c2f395b4395abf5f361b69ce80389ff4aa06fdf0

See more details on using hashes here.

Provenance

The following attestation bundles were made for autoagents_guardrails_py-0.4.0.tar.gz:

Publisher: python-bindings.yml on liquidos-ai/AutoAgents

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file autoagents_guardrails_py-0.4.0-cp39-abi3-manylinux_2_28_x86_64.whl.

File metadata

File hashes

Hashes for autoagents_guardrails_py-0.4.0-cp39-abi3-manylinux_2_28_x86_64.whl
Algorithm Hash digest
SHA256 75fafd88d8300e05226c48eb62ffb376f94077f55bc1657f17ddfa33829ac585
MD5 73c09b2df4163fc9229005ee87412c45
BLAKE2b-256 f2c67074eae49b73816792d88ab37370c3c4360071ebe2f2d272581dd2f75a25

See more details on using hashes here.

Provenance

The following attestation bundles were made for autoagents_guardrails_py-0.4.0-cp39-abi3-manylinux_2_28_x86_64.whl:

Publisher: python-bindings.yml on liquidos-ai/AutoAgents

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file autoagents_guardrails_py-0.4.0-cp39-abi3-manylinux_2_28_aarch64.whl.

File metadata

File hashes

Hashes for autoagents_guardrails_py-0.4.0-cp39-abi3-manylinux_2_28_aarch64.whl
Algorithm Hash digest
SHA256 1f1c2ca2a0569cf8c4e3b1409414e9b7c6c3aa4c2423b0d2c83d5466e137c6c2
MD5 8b9e552ee335f298d5efb21d6adaa9ba
BLAKE2b-256 6140d84ba050386cc8912fa5a4cbcc6a65f0ef87220ef0b95a00abf71a95a104

See more details on using hashes here.

Provenance

The following attestation bundles were made for autoagents_guardrails_py-0.4.0-cp39-abi3-manylinux_2_28_aarch64.whl:

Publisher: python-bindings.yml on liquidos-ai/AutoAgents

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file autoagents_guardrails_py-0.4.0-cp39-abi3-macosx_11_0_arm64.whl.

File metadata

File hashes

Hashes for autoagents_guardrails_py-0.4.0-cp39-abi3-macosx_11_0_arm64.whl
Algorithm Hash digest
SHA256 f968bd2f70df4c2cac76ece5704d7eb6a684fb653c8e362d0f76b85932663d17
MD5 721fcb599ec4caae4307bcdf4696787b
BLAKE2b-256 61b0227b5660a4d722e9a161e253860fcd1a2df688aff00e7be004716bc96ff5

See more details on using hashes here.

Provenance

The following attestation bundles were made for autoagents_guardrails_py-0.4.0-cp39-abi3-macosx_11_0_arm64.whl:

Publisher: python-bindings.yml on liquidos-ai/AutoAgents

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page