2026-02-10 –, From Natural Language to Robot Framework with AI Agents, MCP & Self Healing Things
Turn natural-language scenarios into executable Robot Framework tests and tasks with AI agents, the Model Context Protocol (MCP), and self-healing. In this hands-on workshop, we combine robotframework-aiagent with RF-MCP for contextual assistance and robotframework-selfhealing-agents for resilient execution. Practice AI-assisted test generation and execution, image/document/OCR workflows, and structured extraction. Connect it all in VS Code with RobotCode and GitHub Copilot. Build a multi-model pipeline to generate, run, and heal tests.
The gap between business intent and automated tests and RPA is closing fast. This workshop shows a practical end-to-end path to AI-powered Robot Framework development that translates human-readable scenarios into robust, maintainable test suites.
We start by wiring up robotframework-aiagent for natural-language test execution and AI-assisted test generation. You’ll integrate image analysis, document processing, OCR, and structured data extraction so agents can reason about UI states, PDFs, screenshots, and logs. Next, we introduce the Model Context Protocol (MCP) via RF-MCP to provide agents with rich project context (keywords, resources, env data) and enable semantic keyword matching, interactive step-by-step execution, and state-aware testing with intelligent suggestions and error recovery.
To make your suites resilient in the real world, we bring in robotframework-selfhealing-agents to automatically adapt locators, retry strategies, and flows—so flaky UI changes don’t break your pipeline. You’ll also configure multi-model workflows (OpenAI, Anthropic, Mistral, self-hosted cloud, and local models) and learn when to route tasks to specialized models for token efficiency and quality. In addition, we cover how to test AI applications with non-deterministic input and output using semantic assertions, tolerance windows. Finally, we connect everything inside VS Code with the RobotCode extension’s AI features and GitHub Copilot optimizations—so authoring, refactoring, and debugging are all AI-assisted.
- generate production-ready Robot Framework tests from natural language,
- capture and analyze application state (DOM/API/DB) per step,
- apply self-healing to reduce maintenance,
- convert successful exploratory runs into optimized test suites, and
- run a repeatable, multi-model AI toolchain in CI.
robot ai mcp rpa
Is this suitable for ..?: Beginner RF User, Intermediate RF User, Advanced RF User In-Person or Online talk/workshop?:in person
Many Kasiriha is a QA Engineer at Schenker AG with 16+ years in testing and a board member of the Robot Framework Foundation since 2022. He specializes in test automation training and maintains open source Robot Framework libraries. Based in Düsseldorf, he's a speaker at RoboCon and a father who can't switch off his testing mindset.
Developer, Opensource enthusiast, Pro in many languages and testing tools, Developer of RobotCode -Robot Framework support for VSCode