# Polaxis

> Polaxis is the runtime security and governance layer for autonomous AI agents. It intercepts every tool call before it executes, enforces policy rules, routes high-risk actions to humans for approval, and generates one-click compliance reports.

## What Polaxis Does

Polaxis sits between your AI agents and every tool they call. When an agent tries to send an email, charge a card, delete a record, or call an external API, Polaxis evaluates the action against your policy rules before it executes — and either allows it, blocks it, or pauses it for human approval.

- **Policy engine**: Define rules that block, allow, or escalate any tool call by tool name, parameters, spend amount, or environment
- **Human-in-the-loop approvals**: High-risk actions pause and wait for Slack or email approval before proceeding
- **Immutable audit log**: Every tool call captured with full input payload, policy decision, latency, and threat signals
- **Compliance reports**: One-click PDF for SOC 2 Type II, GDPR, HIPAA, EU AI Act, OWASP Agentic Top 10 — ready in 60 seconds
- **Budget controls**: Per-session and daily spend caps with hard stops
- **PII detection**: Scans tool arguments for credit cards, SSNs, API keys before they leave your system
- **Prompt injection detection**: Catches injection attempts embedded in tool inputs

## Who It's For

Engineering teams and enterprises building and deploying autonomous AI agents — from solo developers shipping their first production agent to enterprise engineering orgs with compliance requirements.

## Framework Support

LangGraph, CrewAI, PydanticAI, OpenAI Agents SDK, LangChain, AutoGen, MCP Protocol, and raw Python agents via `pip install polaxis`.

## Compliance Coverage

- SOC 2 Type II (CC6, CC7, CC8)
- GDPR (Articles 22, 30, 32)
- HIPAA (164.308, 164.312)
- EU AI Act (high-risk AI system requirements)
- OWASP Agentic Top 10 (all 10 categories, full coverage)

## Key Pages

- Homepage: https://polaxis.io
- Blog: https://polaxis.io/blog
- Docs: https://docs.polaxis.io
- Pricing: https://polaxis.io/pricing
- Register: https://polaxis.io/register

## Blog Posts

- How to Audit Your AI Agent: A Complete Guide — https://polaxis.io/blog/how-to-audit-your-ai-agent
- The OWASP Agentic Top 10: What Every AI Team Needs to Know — https://polaxis.io/blog/owasp-agentic-top-10-explained
- The 10 Best AI Agent Security and Governance Tools in 2026 — https://polaxis.io/blog/best-ai-agent-security-governance-tools-2026

## Documentation

- Introduction: https://docs.polaxis.io/introduction
- Quickstart: https://docs.polaxis.io/quickstart
- Policy Engine: https://docs.polaxis.io/policy-engine
- Human Approvals: https://docs.polaxis.io/human-approvals
- Budget Controls: https://docs.polaxis.io/budget-controls
- Compliance Reports: https://docs.polaxis.io/compliance-reports
- Integrations: https://docs.polaxis.io/integrations