Skip to content
O

OpenAPPA

An information-flow policy engine for LLM agents

Open source

Domain Rating by Ahrefs19

#5094in AI
0In stores

Description

OpenAPPA is an information-flow policy engine for LLM agents. It is a deterministic guardrail that tracks how data moves across tool calls,…

Highlights

  • Deterministic policy engine for LLM agent tool calls
  • Tracks data flow across tool calls
  • Ready-made policy batteries, plus your own
  • Works with Archestra, Claude Code, kAgent, Pi and OpenCode
  • MIT licensed, with a playground and benchmarks

Ratings & Reviews

No ratings yet

Be the first to rate OpenAPPA.

People also visit

A

Adrian

Open-source runtime AI agent security tool

AI
Launch
D

DashClaw

Remote approvals, policy checks, and execution evidence for unattended AI agents

AI
Launch
O

Opengeni

Open, self-hostable agentic service for organizations

AI
Launch
L

llm-security

Security guidelines for LLM applications based on OWASP Top 10 for LLM 2025

AI Skills
Get
L

Lacp

Local-first policy, evidence, memory, and recovery controls for Claude, Codex

AI
Launch
A

agent-governance

Declarative policies, intent classification

AI Skills
Get

OpenAPPA FAQ

What is OpenAPPA?

An information-flow policy engine for LLM agents, designed to stop data from leaking through prompt injection or hallucinated tool calls.

How is it different from auto modes?

The site says other guardrails use a second model to judge each tool call and cannot track data flow across calls. OpenAPPA tracks flows instead of matching patterns.

Which agents does it work with?

The site lists Archestra, Claude Code, kAgent, Pi and OpenCode.

What is the license?

MIT. The site also links to a paper and to the GitHub repository.

What are the best OpenAPPA alternatives?

Popular alternatives to OpenAPPA include Adrian, DashClaw and Opengeni. See all OpenAPPA alternatives

Developer Stuff

Badges