Skip to content

herman-prod/docs

Repository files navigation

AARM — Autonomous Action Runtime Management

License: MIT Docs

An open system specification for securing AI-driven actions at runtime.


What is AARM?

As AI systems evolve from text generators into autonomous agents executing consequential actions—API calls, database mutations, financial transactions—the security boundary shifts from model outputs to tool execution.

AARM defines what a runtime security system must do:

  1. Intercept actions before execution
  2. Evaluate against policy
  3. Enforce decisions (allow / deny / modify / require approval)
  4. Record tamper-evident receipts
Agent ──► AARM System ──► Tools/APIs
              │
              ▼
        Action Receipts

Why AARM?

Traditional security doesn't address AI-driven actions:

  • SIEM — observes after execution; can't prevent harm
  • API gateways — verify who, not what the action means
  • Firewalls — agents operate inside the perimeter
  • Prompt guardrails — filter text, not actions
  • Human-in-the-loop — doesn't scale; can be exploited

AARM fills this gap with runtime enforcement at the action layer.


Documentation

aarm.dev


Quick Links

Resource Description
Specification Problem, definition, components
Threat Model What AARM defends against
Architectures Gateway, SDK, eBPF options
Conformance R1–R8 requirements
Research Paper IEEE-style technical paper

Contributing

Contributions welcome. See CONTRIBUTING.md.


Local Development

npm i -g mint    # Install Mintlify CLI
mint dev         # Run at http://localhost:3000

License

MIT


SIEM was built for events. AARM is built for actions.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

Generated from mintlify/starter