analiza aiqbench

This launch is a product release for deploying LLM agents built on an open-source, model-agnostic harness, with configuration for models, instructions, skills, MCP tools, and sandboxes. For AI readers, the main points are portability and control: memory and orchestration stay in the user's hands rather than a single vendor's stack, though the real-world value will depend on production reliability and how well the open components hold up outside the launch demo.