Metadata-Version: 2.4
Name: constrai
Version: 0.0.1
Summary: Formal safety framework for AI agents with provable guarantees
Author-email: Ambar <ambar13@u.nus.edu>
License: MIT
Project-URL: Homepage, https://github.com/Ambar-13/ConstrAI
Project-URL: Repository, https://github.com/Ambar-13/ConstrAI
Project-URL: Bug Tracker, https://github.com/Ambar-13/ConstrAI/issues
Keywords: ai-safety,formal-methods,autonomous-agents,invariants,verification,agent-framework
Classifier: Development Status :: 2 - Pre-Alpha
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: Software Development :: Libraries :: Application Frameworks
Requires-Python: >=3.9
Description-Content-Type: text/markdown

# ConstrAI

**Formal safety framework for AI agents with provable guarantees.**

> Full release coming soon. See [github.com/Ambar-13/ConstrAI](https://github.com/Ambar-13/ConstrAI)

ConstrAI enforces safety at the execution layer, not the prompting layer.
Mathematical invariants, budget control, and tamper-evident audit trails —
provable, not promised.

## Status

This is a name reservation. The full `v0.3.x` release is in active development.
