Metadata-Version: 2.4
Name: enclave42
Version: 0.1.0
Summary: Enclave42 Python client, including the public Confidential inference facade.
Author: Num Lab Tech
Maintainer: Num Lab Tech
License-Expression: Apache-2.0
Project-URL: Homepage, https://enclave42.com
Keywords: confidential-computing,ai-inference,attestation,security,enclave42
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3 :: Only
Classifier: Operating System :: POSIX :: Linux
Classifier: Operating System :: Microsoft :: Windows
Classifier: Operating System :: MacOS
Classifier: Intended Audience :: Developers
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
License-File: THIRD_PARTY_NOTICES.txt
Requires-Dist: cryptography>=43.0
Requires-Dist: pqcrypto==0.3.4
Provides-Extra: verification
Requires-Dist: dcap-qvl==0.6.3; extra == "verification"
Requires-Dist: nv-attestation-sdk; extra == "verification"
Provides-Extra: dev
Requires-Dist: pytest; extra == "dev"
Requires-Dist: pyyaml; extra == "dev"
Requires-Dist: wheel==0.48.0; extra == "dev"
Requires-Dist: pybars3==0.9.7; extra == "dev"
Dynamic: license-file

# enclave42

Python client for Enclave42, including the public Confidential inference facade.

```
pip install enclave42
```

```python
from enclave42 import Enclave42

client = Enclave42(api_key="...")
result = client.confidential.chat(
    model="glm-5-1-tee",
    prompt="Explain this in one sentence.",
)
print(result.text)
print(result.receipt_id)
```

See `examples/confidential_demo.py` in the source repository for a complete example.

`client.confidential.chat(...)` performs exactly one underlying dispatch per call: it creates a
Confidential session, retrieves and locally verifies attestation evidence, encrypts the request
locally, sends it, decrypts the response locally, and returns a `ConfidentialResult` with a real
Security Receipt correlation. It never falls back to an unencrypted request.

Install `enclave42[verification]` to additionally install the dependencies used for stronger,
real TDX/GPU attestation verification. Without them, the corresponding verification dimension(s)
degrade honestly to unsupported/unverified rather than ever being reported as verified.
