Metadata-Version: 2.4
Name: qaisys
Version: 0.1.0a1
Summary: Environment diagnostics for Quantum AI Systems
Author: FlagQuantum Team
License-Expression: Apache-2.0
Classifier: Development Status :: 3 - Alpha
Classifier: Programming Language :: Python :: 3
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Dynamic: license-file

# QAISys

**Quantum AI Systems** tools for inspecting the software environment used by
quantum and AI workloads. The first alpha provides a working environment
report through Python and a command-line interface.

```bash
python -m pip install qaisys==0.1.0a1
qaisys
qaisys --probe-devices
```

```python
from qaisys import diagnose
report = diagnose()
print(report['packages'])
```

The default reads installed distribution metadata for FlagQuantum, FlagQAI,
PyTorch, QSteed, JAX and NumPy. Missing packages appear as `null`. It reports
Python version, operating system and machine architecture, without importing
these optional packages or initializing GPUs.

`--probe-devices` explicitly imports PyTorch and queries CUDA device names and
MPS availability. PyTorch must already be installed to use this option. A failed
probe returns a structured status and exception type, without exception text;
the CLI exits with status 1. Metadata-only inspection exits with status 0 even
when optional packages are missing.

No runtime dependencies; Python 3.10+. Python 3.12 is the tested baseline for
this alpha. Reports describe observations, not compatibility certification,
performance, or a successful quantum workload. Device probing can initialize
drivers; it does not run a benchmark. Output is local JSON and is not uploaded.

## Development

```bash
python -m pip install -e .
python -m unittest discover -s tests -v
```

Apache-2.0 licensed. Public APIs may change during the alpha series.
