Integrations
Wire Context Guardrails into the services that already deliver email and other untrusted context to your agents.
Context Guardrails score web pages, email, files, skills, MCP repositories, and packages before an agent consumes them. Integrations put that scan in the path of a third-party pipeline, so untrusted payloads are scored before your application handles them.
Each guide uses the Context Guardrails API with an organization API key. Create a key in Settings.
Available integrations
| Integration | What it covers |
|---|---|
| Resend | Scan inbound email for phishing, credential harvesting, prompt injection, and social engineering before your application processes it |