provos/ironcurtain
A secure* runtime for autonomous AI agents. Policy from plain-English constitutions. (*https://ironcurtain.dev)
This helps people who want AI agents to automate tasks like managing files or running commands, but need to ensure these agents operate within safe boundaries. It takes plain-English rules, called 'constitutions,' and uses them to control what an AI agent can and cannot do. The result is an AI agent that can work autonomously without accidentally deleting files, exfiltrating data, or pushing malicious code, even if the agent itself tries to go rogue. This is for developers or operations engineers using AI agents who are concerned about security and control.
118 stars. Available on npm.
Use this if you need to deploy autonomous AI agents for tasks like code commits or system interaction, but require strict, clearly defined security policies to prevent unintended or malicious actions.
Not ideal if you're looking for a simple AI agent development framework without a strong emphasis on security policy enforcement or if you need an AI agent that operates without any behavioral constraints.
Stars
118
Forks
18
Language
TypeScript
License
Apache-2.0
Category
Last pushed
Mar 11, 2026
Commits (30d)
0
Dependencies
24
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/agents/provos/ironcurtain"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related agents
DeepTrail/deepsecure
Effortlessly secure your AI agents and AI-powered workflows — from prototype to production. Get...
agentkitai/agentgate
Approval workflows for AI agents
ghostapp-ai/ghost
The Private Agent OS — search files, run AI agents, connect to 10,000+ tools via the complete...
log-bell/avakill
🔪 Open-source safety firewall for AI agents. Intercepts tool calls before they execute, enforces...
node9-ai/node9-proxy
The Execution Security Layer for the Agentic Era. Providing deterministic "Sudo" governance and...