AI & Systems · Knowledge Hub

Eight constellations.
Every skill you need to command coding and enterprise agents.

Eighty-four interactive, fact-checked guides on Claude Code, OpenAI Codex, Microsoft Copilot, Google Gemini CLI, in-editor IDE agents, running them in tandem, the tool-agnostic first principles beneath them all, and the cross-tool toolkit that ties them together — from your first session to orchestrating fleets of agents. Each guide you finish lights a star.

Start with Claude CodeExplore CodexCross the BridgeMap with Gemini CLIEnter the WorkshopLearn First PrinciplesVisit the ObservatoryEnter the Citadel

Hover a constellation to explore · click a star to open its guide

0 / 84 stars explored

The Navigator — Claude CodeThe Forge — CodexThe Bridge — Claude + CodexThe Cartographer — Gemini CLIThe Workshop — IDE AgentsFirst Principles — FoundationsThe Observatory — The ToolkitThe Citadel — Microsoft Copilot

The Navigator

Claude Code

Master the agentic coding loop

Twenty-five guides from first session to orchestrated agent fleets — project memory, context engineering, skills, subagents, hooks, MCP, plugins, the Agent SDK, prompting, test-driven loops, GitHub and CI, enterprise governance, a complete built-in command reference, and the judgment to use them well.

25 of 25 guides live →

The Forge

Codex

OpenAI's coding agent, decoded

Fourteen guides to working with Codex: AGENTS.md, prompting, the full CLI, MCP, cloud and CI automation, the SDK and subagents, code review across every surface, the extension ecosystem, TUI personalization, and an honest comparison with Claude Code.

14 of 14 guides live →

The Bridge

Claude + Codex

Two agents, one repo

Five guides on running Claude Code and Codex in tandem — delegating across tools with the official plugins, bridging them over MCP, splitting plan / execute / review across two models, and orchestrating parallel fan-outs.

5 of 5 guides live →

The Cartographer

Gemini CLI

Google's agent in your terminal

Twelve guides to Google's open-source terminal agent: what Gemini CLI is and how to install and authenticate it, GEMINI.md context, prompting the 1M-token window, tools and slash commands, MCP, extensions, settings, models, sandboxing and approvals, headless automation and CI, and an honest comparison with Claude Code and Codex.

12 of 12 guides live →

The Workshop

IDE Agents

Agents that live in your editor

Eight guides to the AI coding agents embedded in your editor: the 2026 in-editor landscape and agent mode, Cursor essentials and its rules-and-context system, Windsurf/Cascade now Devin Desktop, GitHub Copilot agent and cloud agents, the open-source Cline and Roo Code, AGENTS.md and rules files across tools, and an honest IDE-versus-CLI decision guide.

8 of 8 guides live →

First Principles

Foundations

Craft that outlives any one tool

Seven tool-agnostic guides to the craft beneath every coding agent: the named agentic design patterns, context and prompt engineering, building an MCP server and the protocol behind it, agent security and sandboxing, spec-driven development, and evals that prove your setup actually works.

7 of 7 guides live →

The Observatory

The Toolkit

Tools that see the whole sky

Four cross-tool guides for the experience layer above any single agent: a decision guide that routes each task to the right coding agent, one master matrix comparing every agent in the hub, copy-paste starter kits that make any repo agent-ready, and a shared glossary so each tool’s renaming stops being confusing.

4 of 4 guides live →

The Citadel

Microsoft Copilot

Governed agents for Microsoft 365 work

Nine field-manual guides to the Microsoft Copilot stack: practical daily workflows, the technology map, Copilot Studio, declarative agents and API plugins, MCP tools, Graph connectors, Jira and Confluence automation, enterprise reporting pipelines, and security governance with ALM.

9 of 9 guides live →

Why trust this hub

Researched like a system, not a listicle.

01

Crawled the field

Built from a page-by-page analysis of the leading community guides plus the complete official documentation from Anthropic and OpenAI.

02

Fact-checked every claim

Commands, flags, file paths, model names, and pricing were verified against official sources; anything outdated or wrong was corrected, not copied.

03

Rebuilt to be interactive

Live terminal sessions, config builders, decision trees, and knowledge checks — because you learn agentic coding by doing, not skimming.