Cognitive Architecture Agent

Cognitive Architecture Agent Overview This workflow automates the creation of a comprehensive project specification from a raw, unstructured input. It acts as an AI-powered project manager or technical architect, taking a simple project idea and methodically generating a full suite of project artifacts, metadata, and resource estimations. The entire process is orchestrated to build upon itself, creating richer context for subsequent steps. Technologies Used - n8n: The core orchestrator that connects all services and manages the logic flow. - Supabase: A PostgreSQL database used as the persistent memory for the agent. It stores the initial input, the generated project artifacts, specific prompts for the AI, and the final estimations. - Google Gemini & OpenAI: A suite of AI models (Gemini Pro, Gemini Flash, and a potential gpt-4o-mini for JSON reformatting) are used for the cognitive heavy lifting. - LangChain Nodes: The workflow uses n8n's LangChain integration for advanced AI operations, including chaining prompts, enforcing structured output, and automatically correcting AI errors. What It Helps Automate This workflow automates the tedious and time-consuming initial phase of project planning and specification. It replaces manual work for: - Technical Documentation: Generating summaries, scope definitions, requirements lists, and user stories. - System Design: Creating structured diagrams like Input-Task-Output (ITO) tables and Task-Modality Matrices (TMM). - Project Analysis: Automatically assessing project difficulty and scale based on the generated documents. - Resource Estimation: Calculating project timelines and effort based on a complex set of predefined rules applied to the project's characteristics. By chaining these tasks, it creates a reliable, repeatable, and fast system for turning ideas into actionable project plans. AI Integration AI is the core engine of this workflow, functioning as a multi-agent system where different AI-powered nodes perform specialized tasks. 1.