Shareable instruction files for Cursor, Claude Code, Copilot, Windsurf, and more — installed alongside your existing agent rules.
ArchDesigner employs a sophisticated multi-agent AI system built on advanced Chain-of-Thought (CoT) reasoning patterns. Each agent is a specialized serverless function that performs specific tasks with high accuracy, transparency, and security.
**🤖 This guide helps LLM agents add features to Cumulo applications based on [calcium-workflow](https://github.com/Cumulo/calcium-workflow) template.** 📚 **Related**: [Respo UI](./llms/Respo.md) cr libs readme respo.calcit -f docs/Respo-Agent.md
Skill Master uses the **Claude Agent SDK** to orchestrate an agentic pipeline that generates Claude Agent Skills from unstructured inputs. This document explains the agent architecture, skill generation flow, and how to work with the skills in this project.
Deep technical reference for contributors and AI agents. For coding conventions, standards, and project overview, see [`.github/copilot-instructions.md`](.github/copilot-instructions.md). For user-facing documentation, see [README.md](README.md).
You are an expert autonomous AI developer tasked with building a local Code Intelligence Engine and Model Context Protocol (MCP) server. Your goal is to parse a massive, legacy Python codebase into an Abstract Syntax Tree (AST), convert it into an in-memory Knowledge Graph, and expose tools for AI agents to query.
Always work on tasks longer than you think are needed to accomodate edge case scenarios, testing, ensuring it works, implementing connecting logic and ensuring high quality of standards. NEVER START A PLAN AND THEN ASK FOR USER INPUT, OR ASK FOR USER INPUT DURING A COMPLETION OF A TASK - ONCE YOU ARE ASSIGNED A PROPER TASK - YOU MUST COMPLETE IT END TO END. IF YOU ARE ASSIGNED A TASK OF AN ALREADY COMPLETED TASK THAT YOU ARE ANALYZING TO ENSURE COMPLETENESS, IF YOU DO FIND GAPS DON'T ASK FOR CON