Kaltura, Inc., the Agentic Digital Experience company, today announced the open-source release of a suite of AI agent skills - structured, production-tested knowledge modules that enable AI coding agents such as Claude Code, OpenAI Codex, Google Gravity, GitHub Copilot, and Cursor to build complete applications on the Kaltura platform. Each skill covers a specific Kaltura capability - from AI-powered avatar conversations to video upload and transcoding - and is continuously validated by automated tests running against Kaltura's production APIs, ensuring skills remain always up to date.
Over 100 REST services built from platform inception enable API-first architecture.
Skills include curl examples, parameter tables, error handling patterns, security requirements, and integration best practices.
Kaltura Unisphere provides over 10 embeddable widgets via CDN as ES modules.
Four standards adopted: MCP, Agent Skills Standard, Llms.txt, and Context7 MCP Integration.
Supports client SDKs including Python, JavaScript, PHP, Java, .NET, Ruby, Swift, and Kotlin.
Skills enable avatar-guided portals, intelligent webinars, interactive learning environments, and automated content production pipelines.
AI coding agents are transforming how software is built, but agents are only as capable as the knowledge they can access. Most platform documentation was written for humans browsing web portals - not for AI agents that read, reason, and execute. Each Kaltura skill is a self-contained, machine-readable knowledge module that teaches an AI agent how to accomplish a specific task correctly, including tested curl examples, parameter tables, error handling patterns, security requirements, and integration best practices. When an AI coding agent loads a Kaltura skill, it doesn't just know which endpoints to call - it knows how to build a secure, production-quality integration. Kaltura's API-first architecture is what makes this possible. Every Kaltura capability was designed as a programmable API from day one, allowing each capability to be expressed as a skill an AI agent can learn and execute.
With Kaltura's skills loaded, AI coding agents can build avatar-guided websites, portals and content hubs embedding photorealistic AI avatars into any website or application using the Conversational Avatar SDK, with avatars conducting real-time conversations in 30+ languages with sub-second latency via WebRTC; intelligent webinars and virtual events where Kaltura's Events Platform APIs and dedicated MCP server enable AI coding agents to create, configure, and manage virtual events through natural-language commands; interactive learning environments combining avatar-powered instruction, interactive video quizzes, AI-generated captions and translations in 80+ languages, and conversational AI tutors; and automated content production pipelines generating avatar-narrated videos from scripts, captioning and translating them into 80+ languages, and distributing across channels.
Kaltura has adopted four complementary standards that enable AI coding agents to find, understand, and use the platform autonomously. Model Context Protocol (MCP) provides Kaltura's Events Platform MCP server letting agents search events, create sessions, and manage speakers through natural-language commands. Agent Skills Standard provides structured skill definitions that AI coding agents auto-discover when working in the repository.
About Kaltura
Kaltura (Nasdaq: KLTR) powers rich, agentic digital experiences across organizational journeys for customers, employees, learners, and audiences. Its platform combines intelligent content creation, enterprise-grade content management and intelligence, and multimodal conversational engagement capabilities. Kaltura serves leading enterprises, financial institutions, educational institutions, media and telecom providers, and other organizations worldwide.