agent-designer

Design an AI agent workflow by generating an AGENTS.md blueprint that defines identity, triage tables, sequential phases, and skill activation patterns for robust execution. This structure supports complex workflows requiring triage, 3-6 phased execution steps, and human-in-the-loop integration.

An AGENTS.md blueprint is a production-ready file structure that outlines an AI agent's identity, workflow, triage logic, and skill activation patterns before generating artifacts. It adheres to documented design patterns to ensure robust agent architecture and language rule compliance.

Yes, you can integrate human-in-the-loop phases when building Stratio Cowork agents. The agent architecture design supports 3-6 execution phases, allowing you to map specific stages that require human intervention, progress reporting, and structured interaction guidelines.

Yes, the agent-designer workflow explicitly supports on-demand skill activation patterns. It provides guidance on creating internal and shared skills, defining activation triggers, and integrating them seamlessly into the Stratio Cowork pipeline during the phased execution workflow.

Structure internal and shared skills by defining their activation patterns and integration points within the AGENTS.md blueprint. The design process maps these skills to specific workflow phases, ensuring they trigger on-demand and integrate correctly with the Stratio Cowork pipeline architecture.

You need a triage table in your agent workflow design when handling complex processes that require dynamic routing across 3-6 execution phases. The triage mechanism categorizes inputs to activate the correct sequential phases, skill activations, and human-in-the-loop checkpoints within the Stratio Cowork pipeline.