TASKING has launched the winIDEA AI Kit, a toolkit that lets AI agents integrate with TASKING’s winIDEA development environment to automate embedded software workflows.

The kit consists of three components: the winIDEA MCP Server for structured tool access; AI-ready winIDEA SDK Documentation to help agents generate automation scripts; and Embedded-Development Skills that provide task-oriented guidance for debugging, timing analysis, configuration, and device workflows.

The AI approach is designed to augment engineers’ productivity rather than replace them, with human oversight ensuring accountability and adherence to safety and security requirements.

The kit supports both cloud-based models and locally deployed LLMs, enabling use in environments without internet access.

The kit enables requirements traceability and supports certification to functional safety standards by combining AI automation with human oversight and guardrails to maintain reliability, safety, and security.

TASKING emphasizes that AI serves as a productivity tool to automate complex processes and free engineers from repetitive tasks while maintaining traceability for certification across automotive, aerospace, defense, industrial, medical, and robotics sectors.

AI agents can perform live-debug tasks such as pinpointing crash locations, analyzing CPU load, inspecting variables and memory, configuring traces and code coverage, and executing testing and analysis workflows.

The kit can automate complex tasks across the tool chain, including timing coupling analysis in multicore systems to evaluate worst-case execution timing (WCET) with reduced human error and faster results.

The winIDEA AI Kit is model- and IDE-agnostic, allowing developers to use the LLM and agentic IDE of their choice while maintaining domain-specific tools and data policies.