SDKs & API

Eight Published SDKs. One Governance Engine.

Four first-party SDKs (TypeScript, Python, Go, Java) and four framework adapters — all published to public registries. Install, configure, ship governed AI in minutes.

Current release: v1.1.0

bash

npm install @palveron/sdk-typescript
# or
pip install palveron-sdk

Read the full SDK reference on docs.palveron.com

Language SDKs

First-Class SDKs for Every Stack

Type-safe clients with built-in retry, structured error handling, and automatic trace metadata enrichment. Each package is published to its native registry — no private dependency mirrors required.

TypeScriptv1.1.0npm

@palveron/sdk-typescript

Install

npm install @palveron/sdk-typescript

TypeScript

import { PalveronClient } from "@palveron/sdk-typescript";

const client = new PalveronClient({
  apiKey: process.env.PALVERON_API_KEY,
});

const result = await client.verify({
  prompt: userMessage,
  agent: "support-bot",
});

Quickstart →

Pythonv1.1.0PyPI

palveron-sdk

Install

pip install palveron-sdk

Python

from palveron import PalveronClient

client = PalveronClient(
    api_key=os.environ["PALVERON_API_KEY"],
)

result = client.verify(
    prompt=user_message,
    agent="support-bot",
)

Quickstart →

Gov1.1.0pkg.go.dev

github.com/palveron/palveron-go

Install

go get github.com/palveron/palveron-go

Go

import palveron "github.com/palveron/palveron-go"

client := palveron.NewClient(
    os.Getenv("PALVERON_API_KEY"),
)

result, err := client.Verify(ctx, &palveron.VerifyRequest{
    Prompt: userMessage,
    Agent:  "support-bot",
})

Quickstart →

Javav1.1.0Maven Central

com.palveron:palveron-java

Install

implementation 'com.palveron:palveron-java:1.1.0'

Java

import com.palveron.PalveronClient;

var client = PalveronClient.builder()
    .apiKey(System.getenv("PALVERON_API_KEY"))
    .build();

var result = client.verify(VerifyRequest.builder()
    .prompt(userMessage)
    .agent("support-bot")
    .build());

Quickstart →

Framework Adapters

Native Adapters for AI Frameworks

Drop-in adapters that wrap your existing framework in the Palveron governance pipeline — no application code rewrite needed.

LangChain

Callback handler with automatic trace creation and policy evaluation.

CrewAI

Crew-level governance with per-agent policy assignment.

OpenAI Agents

Drop-in middleware for the OpenAI Agents SDK.

Microsoft AGT

AGT-compatible capability export via toAgtCapabilities().

OpenClaw / MCP

MCP governance package for agent-to-server connections.

No SDK needed? Use the OpenAI-compatible Gateway proxy — just change your base URL.

Deployment

Your Infrastructure, Your Rules

Managed Cloud

Hosted in the EU (Hetzner, Germany). EU data residency by default. Zero maintenance, automatic updates.

Docker Compose

Single-command deployment for development and small teams. Includes PostgreSQL and the gateway.

Helm Charts

Production-grade Kubernetes deployment with horizontal scaling, health checks, and configurable replicas.

Ship Governed AI Today

Install the SDK for your stack, drop in your API key, and every AI request is governed. Quickstart guides at docs.palveron.com.

Start Free docs.palveron.com

Four first-party SDKs

Four framework adapters

OpenAI-compatible proxy fallback