Metadata-Version: 2.4
Name: gondolin
Version: 0.1.0
Summary: Policy-as-Code for AI Agents. Secure your LLM's access to tools with decorators and safe proxies.
Author-email: Jashua Gupta <findjashua@gmail.com>
Project-URL: Homepage, https://github.com/findjashua/gondolin
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Topic :: Security
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Requires-Python: >=3.8
Description-Content-Type: text/markdown
Requires-Dist: PyYAML>=6.0

# Gondolin

**Gondolin** is a security framework for AI Agents. It acts as a "Safe Proxy" layer between your LLM and your sensitive tools (like Jira, Stripe, or AWS).

## The Concept

Instead of giving your Agent raw access to a library, you wrap it in Gondolin. 
Gondolin applies **Policy-as-Code** (YAML + Decorators) to ensure the Agent never executes dangerous actions.
