# Patronus Protect

**Canonical:** https://apis.io/providers/patronus-protect/  
**Website:** https://patronus.studio/  
**APIs profiled:** 4

Patronus Protect, from Casdo Labs GmbH, is an on-device AI firewall that detects prompt injection, PII/DLP, sensitive documents and agentic tool risk in text, public HTTPS webpages, documents and MCP servers before they are used by AI agents, RAG pipelines or LLM applications. It exposes a REST Scan API (control.patronus.studio/api/v1), first-party Rust, TypeScript and Python SDK clients, a local CLI with Codex, Claude Code and DeepSeek plugins, a remote OAuth-protected MCP server and a local stdio MCP server, and the open-source Patronus Ark scanning core.

## Kin Score — 51.6 / 100 (developing)

Scored 2026-10-09 under rubric 0.23.0. Trend: flat (+0.5 from 51.1).

| Facet | Score |
|---|---|
| Discoverability | 76.7 |
| Contract Quality | 52.0 |
| Contract Governance | 18.2 |
| Operational Transparency | 31.6 |
| Developer Ergonomics | 73.8 |
| Access Clarity | 42.1 |

Regulatory layer — **Horizontal (data, software, accessibility, platform)**: 30.0 (matched via fallback).

## Agent readiness — 66.2 (agent-native)

| Dimension | Value |
|---|---|
| Spec Presence | yes |
| Agentic Access | no |
| Reversibility Documented | no |
| MCP Server | verified |
| Auth Clarity | served |
| Idempotency | verified |
| Error Semantics | verified |
| OpenAPI Examples | verified |
| Rate Limit Signal | verified |
| Event Surface Described | no |
| Agent Skills | yes |
| Well Known Catalog | no |
| Consent Identity | no |
| Agent Card | no |
| Dry Run Mode | no |
| Delegated Identity | served |
| Protected Resource Metadata | verified |
| Dynamic Client Registration | yes |
| Agentic Commerce | no |

## APIs (1)

- **Patronus Protect Scan API** — The Scan API from Patronus Protect — 2 operation(s) for scan.

## MCP servers (1)

- **Patronus Protect MCP Server** — Patronus ships an MCP server in two deployments. A remote, OAuth-protected MCP server is served at control.patronus.studio/api/mcp (DCR + PKCE, scopes scan:read/scan:write). A l...

## Security (3)

- **Patronus Protect Authentication** — http/oauth2 · 2 schemes
- **Patronus Protect Domain Security** — TLSv1.3 · DMARC
- **Patronus Protect Vulnerability Disclosure** — security.txt · contact published

## Plans (1)

- **Patronus Protect Plans Pricing**

## Tags

AI Safety, Prompt Injection, Security, LLM, Agent-Native, MCP, DLP, PII, On-Device, AI Firewall

---

Profiled by [API Evangelist](https://apievangelist.com) and published on [APIs.io](https://apis.io/providers/patronus-protect/). Scores are computed from the provider's own public artifacts under a published rubric.