Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions README.md
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
![Agent Kernel Banner](docs/static/img/banner.png)

# Agent Kernel by Yaala Labs
<p align="center">
<img src="docs/static/img/branding/agent-kernel-lockup-dark.svg" alt="Agent Kernel" width="420" />
</p>

> **[Developer Guide](DEVELOPER_GUIDE.md)**

Expand Down
36 changes: 22 additions & 14 deletions docs/docs/intro.md
Original file line number Diff line number Diff line change
Expand Up @@ -5,26 +5,26 @@ slug: /

# Introduction to Agent Kernel

Welcome to **Agent Kernel** - a versatile, framework-agnostic runtime for building and deploying AI agents.
**From Agent Logic to Production in Minutes.**

:::tip What's New
☁️ **Multi-Cloud Azure Support** - Agent Kernel now supports Microsoft Azure! Deploy to Azure Functions (serverless) or Azure Container Apps (containerized) with the same ease as AWS. Full Terraform modules, Cosmos DB session storage, and enterprise-ready deployment patterns. [Learn more →](/docs/deployment/overview)
:::

## What is Agent Kernel?

Agent Kernel is a **lightweight, multi-cloud AI agent runtime** and adapter layer for building and running AI agents across multiple frameworks and cloud providers. It provides the low level scaffolding to build, test and deploy your agents, your MCP tools and A2A quickly in many deployment configurations on **AWS and Azure**. The unified execution environment provides the session and memory management seamlessly.
Agent Kernel is an **open-source runtime** that lets you build, test, and deploy AI agents to production in days instead of months. It works with any major AI framework — OpenAI, LangGraph, CrewAI, Google ADK — and can run agents from multiple frameworks together in a single runtime. It deploys to AWS, Azure, or your own servers with zero platform code. Built-in integrations for Slack, WhatsApp, and more mean your agents can reach users instantly.

**Supported Python Versions:** 3.12 - 3.13.x
**Supported Cloud Platforms:** AWS, Azure
**Think of it like Express.js for web servers, or Spring Boot for Java microservices** — but for AI agents. It gives you the scaffolding, execution environment, session management, and deployment infrastructure so you can focus on writing the logic that matters.

Migrate your existing agents to Agent Kernel and instantly utilize pre-built execution and testing capabilities. It eliminates the complexity of framework development allowing AI engineers to focus on Agent development and provides a consistent development experience regardless of the underlying AI agent framework.
**Supported Python Versions:** 3.12 - 3.13.x
**Supported Cloud Platforms:** AWS, Azure

It's not
- a substitute for popular Agent frameworks and SDKs like LangGraph and OpenAI
It's not:
- a substitute for popular agent frameworks and SDKs like LangGraph, OpenAI Agents, CrewAI, or Google ADK
- another heavy abstraction that you have to learn

It's a lightweight, simple, intuitive framework to make your life easy.
It's a **lightweight, thin adapter** that wraps your existing agents and instantly provides everything else — testing, deployment, integrations, session management, observability.

```mermaid
---
Expand Down Expand Up @@ -235,13 +235,21 @@ All without changing your agent code!

## Who Should Use Agent Kernel?

Agent Kernel is ideal for:
Agent Kernel is built for four types of teams:

- **AI Engineers** who want framework flexibility without vendor lock-in
- **Teams** building production AI agent systems across multiple clouds
- **Developers** who need to migrate between frameworks or cloud providers
- **Organizations** requiring enterprise-grade agent deployment with multi-cloud strategy
- **Researchers** exploring different agent frameworks
### Software Companies (Services)
Development houses and IT services firms with clients asking for AI-powered solutions. Agent Kernel lets them stand up AI agent capabilities quickly without a 6-month R&D cycle — focusing developer time on client-specific agent logic, not infrastructure.

### Software Companies (Products)
SaaS and enterprise software companies wanting to embed AI agents into their products. Agent Kernel's framework-agnostic design eliminates lock-in — add intelligent agents today, switch frameworks tomorrow, and your platform code never changes.

### AI Startups
Early to growth-stage startups building AI-native products. With open-source, no licensing costs, and full deployment infrastructure (Terraform, Docker) out of the box, startups go from prototype to production in days — not quarters.

### Domain Experts
Subject matter experts in finance, healthcare, legal, education, or other fields who want to build AI products without a fulltime engineering team. Agent Kernel dramatically reduces the software engineering surface area — define your agent logic, and Agent Kernel handles the rest.

→ **[Explore all use cases →](/use-cases)**

## Next Steps

Expand Down
64 changes: 38 additions & 26 deletions docs/docusaurus.config.js
Original file line number Diff line number Diff line change
Expand Up @@ -9,8 +9,8 @@ import { themes as prismThemes } from 'prism-react-renderer';
/** @type {import('@docusaurus/types').Config} */
const config = {
title: 'Agent Kernel',
tagline: 'Multi-cloud production-ready framework-agnostic runtime for AI agents - Deploy agents built with your favorite framework to AWS or Azure with enterprise features',
favicon: 'img/favicon.ico',
tagline: 'From Agent Logic to Production in Minutes.',
favicon: 'img/branding/agent-kernel-icon-color.svg',
url: 'https://kernel.yaala.ai',
baseUrl: '/',

Expand All @@ -19,6 +19,8 @@ const config = {
projectName: 'agent-kernel',
trailingSlash: false,

clientModules: ['./src/clientModules/scrollReveal.ts'],

onBrokenLinks: 'throw',

// SEO head tags
Expand All @@ -27,7 +29,7 @@ const config = {
tagName: 'meta',
attributes: {
name: 'keywords',
content: 'agent kernel, agentic framework, agentic frameworks, AI agent runtime, LangGraph runtime, OpenAI Agents deployment, Google ADK deployment, CrewAI deployment, agent kernel vs semantic kernel, production AI agents, AI agent orchestration, multi-agent systems, LangGraph production, OpenAI Agents SDK, Google Gemini ADK, framework-agnostic agents, agent deployment platform, stateful AI agents, AI agent monitoring, LangGraph alternative, OpenAI agent framework, Python AI agents, enterprise AI agents, AI DevOps, MLOps agents, agent-to-agent communication, serverless agents, containerized agents',
content: 'agent kernel, AI agent framework, AI agent runtime, AI agent deployment, AI agent platform, agentic AI development, multi-cloud AI agents, framework-agnostic AI, deploy AI agents, how to build AI agents, how to deploy AI agents to AWS, how to deploy AI agents to Azure, AI agent development platform open source, build AI agents with Python, LangGraph vs OpenAI agents vs CrewAI, AI agent Slack integration, WhatsApp AI bot, production AI agent deployment, AI agent testing framework, AI agent session management, AI agent guardrails, AI agent safety, agent-to-agent communication, MCP protocol, model context protocol, AI agent observability, AI agent tracing, agentkernel, Yaala Labs, LangGraph runtime, OpenAI Agents SDK deployment, Google ADK deployment, CrewAI deployment, serverless AI agents, containerized AI agents, multi-agent systems, AI agent orchestration, enterprise AI agents, Python AI agents, open source AI agent platform',
},
},
{
Expand All @@ -41,14 +43,14 @@ const config = {
tagName: 'meta',
attributes: {
property: 'og:title',
content: 'Agent Kernel - Production-ready framework-agnostic runtime for AI agents',
content: 'Agent Kernel — From Agent Logic to Production in Minutes',
},
},
{
tagName: 'meta',
attributes: {
property: 'og:description',
content: 'Agent Kernel: The production-ready agentic framework-agnostic runtime. Deploy LangGraph, OpenAI Agents SDK, Google ADK, and CrewAI with built-in state management, monitoring, and enterprise scalability.',
content: 'Agent Kernel is an open-source runtime that lets you build, test, and deploy AI agents to production in days instead of months. Works with any major AI framework — OpenAI, LangGraph, CrewAI, Google ADKand can run agents from multiple frameworks together in a single runtime. Deploys to AWS, Azure, or your own servers with zero platform code.',
},
},
{
Expand Down Expand Up @@ -83,14 +85,14 @@ const config = {
tagName: 'meta',
attributes: {
name: 'twitter:title',
content: 'Agent Kernel - Production-ready framework-agnostic runtime for AI agents',
content: 'Agent Kernel — From Agent Logic to Production in Minutes',
},
},
{
tagName: 'meta',
attributes: {
name: 'twitter:description',
content: 'Production agentic framework runtime for LangGraph, OpenAI Agents SDK, Google ADK, and CrewAI. Enterprise-ready framework with state management and monitoring.',
content: 'Open-source, framework-agnostic, multi-cloud AI agent runtime. Build, test, and deploy production AI agents to AWS or Azure in days — not months. Works with OpenAI, LangGraph, CrewAI, and Google ADK.',
},
},
{
Expand Down Expand Up @@ -153,7 +155,7 @@ const config = {
'@type': 'SoftwareApplication',
name: 'Agent Kernel',
applicationCategory: 'DeveloperApplication',
description: 'Production-ready agentic framework runtime for LangGraph, OpenAI Agents SDK, Google ADK, and CrewAI. Framework-agnostic runtime.',
description: 'Agent Kernel is an open-source, framework-agnostic, multi-cloud runtime for production AI agents. Build, test, and deploy AI agents to AWS or Azure in days instead of months. Works with OpenAI, LangGraph, CrewAI, and Google ADK — run multiple frameworks simultaneously in a single runtime.',
operatingSystem: 'Linux, macOS, Windows',
offers: {
'@type': 'Offer',
Expand All @@ -166,16 +168,20 @@ const config = {
url: 'https://www.yaalalabs.com',
},
url: 'https://kernel.yaala.ai',
keywords: 'Agent kernel, agentic framework, LangGraph runtime, OpenAI Agents, Google ADK, CrewAI, AI agent deployment, agent kernel vs semantic kernel, Agentic AI',
keywords: 'Agent Kernel, agentkernel, Yaala Labs, AI agent framework, AI agent runtime, AI agent deployment, agentic AI, multi-cloud AI agents, framework-agnostic AI, LangGraph, OpenAI Agents, CrewAI, Google ADK',
featureList: [
'LangGraph support',
'OpenAI Agents SDK support',
'Google ADK support',
'CrewAI support',
'Multi-agent orchestration',
'State management',
'Production monitoring',
'Framework-agnostic runtime',
'Framework-agnostic runtime — OpenAI Agents, LangGraph, CrewAI, Google ADK',
'Multi-framework simultaneous execution in a single runtime',
'Multi-cloud deployment — AWS Lambda, ECS, Azure Functions, Container Apps',
'Built-in Slack, WhatsApp, Messenger, Instagram, Telegram, Gmail integrations',
'Session and memory management — Redis, DynamoDB, Cosmos DB, in-memory',
'Execution hooks for guardrails, RAG, analytics, and moderation',
'pytest-integrated agent testing framework',
'OpenAI and AWS Bedrock guardrails for content safety',
'LangFuse and OpenLLMetry observability',
'MCP Server and A2A Server protocol support',
'Full Terraform modules for AWS and Azure',
'Open-source — Apache 2.0 license',
],
}),
},
Expand All @@ -193,7 +199,7 @@ const config = {
sameAs: [
'https://github.com/yaalalabs/agent-kernel',
],
description: 'Enterprise AI infrastructure and agentic framework solutions',
description: 'Yaala Labs is a technology company that builds next-generation cloud-native marketplace and capital markets infrastructure for both traditional and digital assets. Creator and maintainer of Agent Kernel, the open-source, framework-agnostic, multi-cloud runtime for production AI agents.',
}),
},
],
Expand Down Expand Up @@ -267,11 +273,12 @@ const config = {
const items = await defaultCreateSitemapItems(rest);
return items.map((item) => {
// Set highest priority for key landing pages
if (item.url === 'https://kernel.yaala.ai/' ||
if (item.url === 'https://kernel.yaala.ai/' ||
item.url === 'https://kernel.yaala.ai/docs' ||
item.url === 'https://kernel.yaala.ai/docs/quick-start' ||
item.url === 'https://kernel.yaala.ai/docs/installation' ||
item.url === 'https://kernel.yaala.ai/features' ||
item.url === 'https://kernel.yaala.ai/use-cases' ||
item.url === 'https://kernel.yaala.ai/blog') {
return {...item, priority: 1.0, changefreq: 'daily'};
}
Expand Down Expand Up @@ -300,9 +307,9 @@ const config = {
({
// SEO metadata
metadata: [
{
{
name: 'description',
content: 'Agent Kernel: Production-ready agentic framework runtime for LangGraph, OpenAI Agents SDK, Google ADK, and CrewAI. Deploy and scale AI agents with enterprise features. Framework-agnostic framework with built-in state management, monitoring, and observability.',
content: 'Agent Kernel: The open-source, framework-agnostic, multi-cloud runtime for production AI agents. Build, test, and deploy with OpenAI, LangGraph, CrewAI, or Google ADK to AWS or Azure — in days, not months. Built-in messaging integrations, session management, testing framework, guardrails, and observability.',
},
{
property: 'og:image',
Expand All @@ -318,10 +325,11 @@ const config = {
title: 'Agent Kernel',
logo: {
alt: 'Agent Kernel Logo',
src: 'img/logo.svg',
src: 'img/branding/agent-kernel-icon-color.svg',
},
items: [
{ to: '/features', label: 'Explore Features', position: 'left' },
{ to: '/features', label: 'Features', position: 'left' },
{ to: '/use-cases', label: 'Use Cases', position: 'left' },
{
type: 'docSidebar',
sidebarId: 'tutorialSidebar',
Expand Down Expand Up @@ -377,6 +385,10 @@ const config = {
label: 'Getting Started',
to: '/docs',
},
{
label: 'Use Cases',
to: '/use-cases',
},
{
label: 'Architecture',
to: '/docs/architecture/overview',
Expand Down Expand Up @@ -447,9 +459,9 @@ const config = {
additionalLanguages: ['python', 'bash', 'json', 'yaml'],
},
colorMode: {
defaultMode: 'light',
disableSwitch: false,
respectPrefersColorScheme: true,
defaultMode: 'dark',
disableSwitch: true,
respectPrefersColorScheme: false,
},
mermaid: {
options: {
Expand Down
59 changes: 59 additions & 0 deletions docs/src/clientModules/scrollReveal.ts
Original file line number Diff line number Diff line change
@@ -0,0 +1,59 @@
let activeObserver: IntersectionObserver | null = null;

function initScrollReveal() {
if (activeObserver) {
activeObserver.disconnect();
activeObserver = null;
}

// Target section elements that are direct children of <main>
// This matches custom pages (homepage, features, use-cases) but NOT docs pages
// (docs pages use <div> children inside <main>, not <section>)
const sections = Array.from(document.querySelectorAll('main > section'));
if (!sections.length) return;

// Respect reduced-motion preference
if (window.matchMedia('(prefers-reduced-motion: reduce)').matches) {
sections.forEach((s) => s.classList.add('scroll-reveal-visible'));
return;
}

// Count sections already in the viewport for stagger timing
let inViewIndex = 0;

activeObserver = new IntersectionObserver(
(entries) => {
entries.forEach((entry) => {
if (entry.isIntersecting) {
const delay = Number(entry.target.dataset.revealDelay ?? 0);
setTimeout(() => {
entry.target.classList.add('scroll-reveal-visible');
}, delay);
activeObserver.unobserve(entry.target);
}
});
},
{ threshold: 0.06, rootMargin: '0px 0px -40px 0px' }
);

sections.forEach((section) => {
section.classList.remove('scroll-reveal-visible');
section.classList.add('scroll-reveal');

const rect = section.getBoundingClientRect();
if (rect.top < window.innerHeight) {
// Section already visible on load — stagger with 80ms between each
section.dataset.revealDelay = String(inViewIndex * 80);
inViewIndex++;
} else {
section.dataset.revealDelay = '0';
}

activeObserver.observe(section);
});
}

export function onRouteDidUpdate() {
// Double rAF ensures React has finished painting the new route
requestAnimationFrame(() => requestAnimationFrame(initScrollReveal));
}
Loading
Loading