Metadata-Version: 2.4
Name: diffusiongemma-agent
Version: 0.1.0
Summary: Installer and CLI for the local DiffusionGemma coding agent runtime
Author: DiffusionGemma Agent contributors
License-Expression: Apache-2.0
Project-URL: Documentation, https://huggingface.co/aogavrilov/diffusiongemma-agent-iq3-cuda13
Project-URL: Runtime, https://huggingface.co/aogavrilov/diffusiongemma-agent-iq3-cuda13
Keywords: diffusiongemma,gguf,coding-agent,wsl,cuda
Classifier: Development Status :: 3 - Alpha
Classifier: Environment :: Console
Classifier: Operating System :: Microsoft :: Windows :: Windows 10
Classifier: Operating System :: Microsoft :: Windows :: Windows 11
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Software Development :: Code Generators
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: huggingface-hub<2,>=1.0
Provides-Extra: publish
Requires-Dist: build>=1.2; extra == "publish"
Requires-Dist: twine>=6; extra == "publish"
Provides-Extra: test
Requires-Dist: pytest>=8; extra == "test"
Dynamic: license-file

# DiffusionGemma Agent

Small installer and command-line client for the Windows + WSL2 full-GPU
DiffusionGemma coding-agent runtime. Model weights and CUDA runtime files are
downloaded separately from a versioned Hugging Face model repository.

```powershell
python -m pip install diffusiongemma-agent
dg-agent install --accept-licenses
dg-agent run --repo C:\work\repo --task "Fix src/x.py and run tests" --file src/x.py
```

The equivalent launcher-independent form is `python -m diffusiongemma_agent`.

Requirements: Windows 10/11, initialized WSL2 Ubuntu, a current NVIDIA Windows
driver with WSL CUDA support, around 16 GB VRAM, and sufficient disk space for
the 12+ GiB IQ3 runtime.

The installer does not enable native model tool calls. Repository changes go
through the checkpointed retrieval, Aider, verification, and rollback path.
