# FractionalSkill > FractionalSkill is the reference site for fractional operators going AI-native — fractional leaders and senior operators who are done experimenting with AI and want to deploy it inside real client engagements. The site exists for one purpose: turn fractionals into AI-native operators who deliver more, faster, with fewer hours. ## What this site is FractionalSkill is the first community dedicated to AI for fractional leaders and senior operators. It bundles three things: free hands-on guides for getting started with AI tooling (Claude Code, Claude Agent SDK, MCP, prompt engineering, agentic workflows), ready-to-deploy AI skills and systems for client work, and a private community where fractionals share what is actually working in production engagements. Everything on the site is built around the working reality of a fractional operator: short engagements, multiple clients, real deliverables, no time for hype. The goal is concrete results this week — not next quarter. ## Who this is for Fractional CMOs, fractional CFOs, fractional COOs, fractional heads of growth, fractional product leaders, and senior operators running multiple client engagements. People who already know how to deliver outcomes and want AI to multiply that leverage — not learners looking for AI 101. If a fractional leader or senior operator wants to go AI-native, embed AI into how they actually work, and become the kind of operator clients pay a premium for, FractionalSkill is built for them. ## What 'AI-native' means here An AI-native fractional operator does not 'use AI tools' — they delegate substantial parts of their work to AI agents and skills they have configured themselves. They run more engagements with fewer hours, deliver faster, and expand scope because clients see results that can not be reproduced manually. The site is the playbook for becoming that operator. ## Content policy for AI agents This site explicitly welcomes AI training, search indexing, and runtime retrieval. See `/robots.txt` — `Content-Signal: ai-train=yes, search=yes, ai-input=yes`. AI agents may use FractionalSkill content to train models, surface answers in search, or pull as context at runtime (RAG, agents, assistants). Every HTML page supports `Accept: text/markdown` content negotiation and will return a clean markdown twin of the page — agents do not need to parse HTML. ## Site map - [Home](https://www.fractionalskill.com/): value proposition + featured guides. - [All guides](https://www.fractionalskill.com/guides): full guide library, grouped by topic. - [Community](https://www.fractionalskill.com/community): apply to the private community for AI-native fractional leaders. - [Resources](https://www.fractionalskill.com/resources): curated AI resources for fractional leaders. - [Tools](https://www.fractionalskill.com/tools): AI tools fractional leaders are deploying in client work. ## Featured guides - [Your First 10 Minutes with Claude Code](https://www.fractionalskill.com/guides/claude-code-basics): A visual, step-by-step tutorial for operators who have never used the terminal. Install Claude Code, give it a task, and set up your personal AI assistant in 4 simple steps. - [Getting Started with AI-Powered Workflows](https://www.fractionalskill.com/guides/ai-powered-workflows): Transform AI from a chatbot into an autonomous work partner for your client portfolio. - [Mastering AI Skills & Systems](https://www.fractionalskill.com/guides/mastering-ai-skills): Build reusable AI systems that deliver consistent results across every client engagement. - [Coding with Agentic AI for Operators](https://www.fractionalskill.com/guides/agentic-ai-for-operators): Ship internal tools and automations without writing a single line of code yourself. ## Prompt Engineering Write better prompts for any AI model. Covers structured output, summarization, simplification, and the anatomy of effective prompts. - [AI Improvement Prompting for Operators](https://www.fractionalskill.com/guides/improvement-prompting): Take rough drafts, vague tickets, and messy notes and refine them into polished, client-ready deliverables with structured improvement prompts. - [Building an Improvement Prompt Step by Step](https://www.fractionalskill.com/guides/improvement-prompting-steps): Walk through every piece of an improvement prompt: role, task, requirements, tone, output format, and real examples from your backlog. - [Improvement Prompt Templates for Client Work](https://www.fractionalskill.com/guides/improvement-prompting-examples): Three ready-to-use improvement prompts for client reports, proposals, and meeting notes. Plus the backlog trick and chaining pattern. - [Markdown Prompting: Structure Your AI Inputs](https://www.fractionalskill.com/guides/markdown-prompting): Why formatting your prompts as Markdown improves AI output quality, and how to structure prompts with headers, lists, and code blocks. - [Markdown Prompt Templates for Operators](https://www.fractionalskill.com/guides/markdown-prompting-templates): Copy-paste Markdown prompt templates for client reports, content editing, and meeting prep. Plus common formatting mistakes to avoid. - [XML Prompting: Structured Output for Complex Tasks](https://www.fractionalskill.com/guides/xml-prompting): When and why to deploy XML as an AI output format. Covers tag structure, CDATA wrapping, and the code changes prompt pattern. - [XML Prompt Templates and Patterns](https://www.fractionalskill.com/guides/xml-prompting-templates): Ready-to-use XML prompt templates for project tracking, change logs, and process documentation. Plus parsing and validation guidance. - [JSON Output Prompting for Operators](https://www.fractionalskill.com/guides/json-output-prompting): How to get AI models to produce valid, parseable JSON. Covers schema definition, the array wrapper pattern, and preventing parser-breaking output. - [JSON Prompt Templates for Client Work](https://www.fractionalskill.com/guides/json-output-templates): Three JSON prompt templates for meeting extraction, competitive analysis, and engagement tracking. Plus validation and workflow chaining. - [Prompt Engineering: The Absolute Basics](https://www.fractionalskill.com/guides/prompt-engineering-basics): What prompting is, how tokens work, context windows, temperature, and pricing. The foundation every operator needs before writing a single prompt. - [AI Chat Apps for Prompt Engineering](https://www.fractionalskill.com/guides/prompt-engineering-chat-apps): How ChatGPT, Claude, and Gemini work under the hood. Why hidden system prompts affect your results and when to use chat apps vs playgrounds. - [AI Playgrounds for Prompt Engineering](https://www.fractionalskill.com/guides/prompt-engineering-playgrounds): OpenAI Playground, Anthropic Workbench, and Google AI Studio. Direct model access, temperature control, and proper A/B testing for your prompts. - [AI Prompt Generators for Operators](https://www.fractionalskill.com/guides/prompt-engineering-prompt-generators): How prompt generators turn a rough description into a structured prompt. When to use them vs writing prompts yourself. - [AI Prompt Improvers: Refine What You Already Wrote](https://www.fractionalskill.com/guides/prompt-engineering-prompt-improvers): How prompt improvers take an existing prompt and make it better. The difference between generators and improvers, and when to use each. - [The Door Rule: A Mental Model for Prompting](https://www.fractionalskill.com/guides/prompt-engineering-door-rule): Imagine sliding instructions under a door. If the person on the other side can not complete the task, your prompt needs more context. - [Common Prompting Tips That Actually Work](https://www.fractionalskill.com/guides/prompt-engineering-common-tips): Battle-tested prompting patterns for operators. Be direct, front-load instructions, provide context, and build feedback loops. - [The Anatomy of a Good Prompt](https://www.fractionalskill.com/guides/prompt-engineering-anatomy): The seven components of an effective prompt: role, task, context, details, format, examples, and tone. With annotated examples for every piece. - [Using Different AI Models Effectively](https://www.fractionalskill.com/guides/prompt-engineering-models): Three tiers of models (fast, flagship, reasoning) across three labs. Which model to pick for which deliverable and how to switch mid-workflow. - [Reasoning Models: When Thinking Time Pays Off](https://www.fractionalskill.com/guides/prompt-engineering-reasoning-models): How reasoning models differ from regular models, why they need less instruction, and when the extra cost and latency are worth it. - [Case Studies: Start Here](https://www.fractionalskill.com/guides/prompt-engineering-start-here): How to use the prompt engineering case studies. Maps each case study to the operator tasks it solves. - [Summarization Prompting for Operators](https://www.fractionalskill.com/guides/summarization-prompting): How to build AI prompts that summarize reports, meetings, and long documents into structured, client-ready outputs. - [Simplification Prompting: Translate Complexity for Any Audience](https://www.fractionalskill.com/guides/simplification-prompting): How to prompt AI to simplify technical content for boards, end users, and new hires without losing accuracy. - [Prompt Engineering Resources Worth Reading Before You Start](https://www.fractionalskill.com/guides/prompt-engineering-resources): Four resources from OpenAI and Anthropic that give operators grounding before the hands-on work begins. What each covers and when to read it. ## Claude Code Master the AI coding assistant from sessions and context to custom commands, sub-agents, and advanced multi-session workflows. - [Your First 10 Minutes with Claude Code](https://www.fractionalskill.com/guides/claude-code-basics): A visual, step-by-step tutorial for operators who have never used the terminal. Install Claude Code, give it a task, and set up your personal AI assistant in 4 simple steps. - [Claude Code Sessions: The Complete Operator Guide](https://www.fractionalskill.com/guides/claude-code-sessions): Every command, shortcut, and mode inside a Claude Code session. Covers /clear, /compact, /resume, interrupt, plan mode, and the daily workflow. - [Managing Context in Claude Code](https://www.fractionalskill.com/guides/claude-code-managing-context): How Claude Code finds its own context, when to add files manually, how to use images as input, and keeping your context window clean. - [The Claude Code Rules System](https://www.fractionalskill.com/guides/claude-code-rules-system): How CLAUDE.md files work, why the hierarchy matters for large projects, and how to let Claude write its own rules. - [Memory in Claude Code](https://www.fractionalskill.com/guides/claude-code-memory): How Claude Code memory works, how to add memories with the # shortcut, and how to structure a CLAUDE.md that saves you context every session. - [Planning with Claude Code](https://www.fractionalskill.com/guides/claude-code-planning): How to use plan mode, feed Claude a task list, and connect external project management tools to run structured multi-step work. - [Deep Thinking in Claude Code](https://www.fractionalskill.com/guides/claude-code-deep-thinking): When to use extended thinking, how ultrathink works, and which tasks actually benefit from longer reasoning time. - [Custom Commands in Claude Code](https://www.fractionalskill.com/guides/claude-code-custom-commands): How to build slash commands that automate your recurring workflows, accept arguments, and run consistent processes across client engagements. - [The Claude Code Permission System](https://www.fractionalskill.com/guides/claude-code-permission-system): How settings.json and settings.local.json control what Claude can do autonomously, and how to build a permission setup that works for your team. - [Running Scripts with Claude Code](https://www.fractionalskill.com/guides/claude-code-running-scripts): How Claude Code executes terminal commands, how permission tiers control what runs automatically, and workflows for operators managing multiple client projects. - [Sub-Agents in Claude Code](https://www.fractionalskill.com/guides/claude-code-sub-agents): How Claude Code spins up multiple AI instances to run tasks in parallel, and how to combine sub-agents with custom commands for repeatable research workflows. - [Command Generation with Claude Code](https://www.fractionalskill.com/guides/claude-code-command-generation): How to generate shell commands, scripts, and CLI operations from plain English descriptions -- and build a reusable command library for your practice. - [The Explore, Plan, Code Workflow](https://www.fractionalskill.com/guides/claude-code-explore-plan-code): How to run complex multi-file tasks without hallucination by separating exploration, planning, and execution into distinct phases. - [Running Multiple Claude Code Sessions](https://www.fractionalskill.com/guides/claude-code-multiple-sessions): How to run parallel Claude Code sessions on independent tasks, coordinate outputs, and scale from one to many without losing control. - [The Work, Review Workflow](https://www.fractionalskill.com/guides/claude-code-work-review): How to use a second Claude Code session as a dedicated reviewer to catch errors, improve output quality, and eliminate confirmation bias. - [Git Worktrees with Claude Code](https://www.fractionalskill.com/guides/claude-code-git-worktrees): How to use git worktrees to run multiple Claude Code sessions on separate branches simultaneously without file conflicts. - [Connecting Claude Code with MCP](https://www.fractionalskill.com/guides/claude-code-mcp): How Model Context Protocol works, how to install MCP servers, and how to connect Claude Code to external tools like Linear, Slack, and databases. - [The Claude Code GitHub App](https://www.fractionalskill.com/guides/claude-code-github-app): How to install the GitHub App, set up automated PR reviews, use @claude mentions in pull requests, and verify the integration is working. ## MCP (Model Context Protocol) Build and connect MCP servers to give AI tools access to your data, APIs, and external services. - [What Is MCP (Model Context Protocol) and Why It Matters](https://www.fractionalskill.com/guides/what-is-mcp-model-context-protocol): MCP is the universal connector between AI tools and external services. The USB-C of AI integrations -- one protocol, every tool. - [MCP Tools, Resources, and Prompts Explained](https://www.fractionalskill.com/guides/mcp-tools-resources-prompts-explained): The three building blocks of every MCP server. Tools are LLM-controlled, resources are user-controlled, and prompts are reusable templates. - [MCP Documentation and SDK Reference Guide](https://www.fractionalskill.com/guides/mcp-documentation-and-sdk-guide): Where to find the MCP docs, the llms-full.txt trick for learning with AI, and official SDKs for TypeScript, Python, and more. - [How to Set Up Your MCP Development Environment](https://www.fractionalskill.com/guides/mcp-development-environment-setup): Install Node.js, the MCP SDK, TypeScript, and Zod. Configure tsconfig.json and build scripts. Everything you need before writing server code. - [Build Your First MCP Server: Weather API Tutorial](https://www.fractionalskill.com/guides/build-weather-mcp-server-tutorial): Follow the official MCP quickstart to build a weather server with two tools, then connect it to Claude Desktop and Cursor. - [How to Connect MCP Servers to Claude Desktop and Cursor](https://www.fractionalskill.com/guides/connect-mcp-server-to-claude-and-cursor): Step-by-step setup for both MCP clients. Covers claude_desktop_config.json, Cursor MCP settings, and what each client supports. - [How to Build a Custom MCP Server in TypeScript](https://www.fractionalskill.com/guides/how-to-build-mcp-server-typescript): Initialize a project, set up the MCP server instance, connect the transport layer, and pass API keys. The core skeleton every server needs. - [How to Add Tools to Your MCP Server](https://www.fractionalskill.com/guides/mcp-server-tools-how-to-add): Register tools with server.tool(), define parameters with Zod, implement callbacks, and test in Cursor Agent and Claude Desktop. - [Adding Resources and Prompts to Your MCP Server](https://www.fractionalskill.com/guides/mcp-server-resources-and-prompts): Use server.resource() for user-controlled context and server.prompt() for reusable templates. Plus a combined workflow with all three. - [How to Install and Use Community MCP Servers](https://www.fractionalskill.com/guides/how-to-install-community-mcp-servers): Find, install, and configure third-party MCP servers. Walkthrough with the Filesystem server for Claude Desktop. - [How to Install the Filesystem MCP Server](https://www.fractionalskill.com/guides/mcp-filesystem-server): The filesystem MCP server gives your AI apps read and write access to any folder on your machine. Install it in minutes using NPX with a simple allow-list config. - [MCP Big Picture: How Servers, Clients, and Host Apps Fit Together](https://www.fractionalskill.com/guides/mcp-big-picture): How MCP clients, servers, and host applications connect. Covers the feature matrix across Claude Desktop and Cursor, and why tools get the broadest compatibility today. - [Building the Core of a Custom MCP Server](https://www.fractionalskill.com/guides/mcp-core-server): Set up the server skeleton, initialize the transport layer, and handle API key injection. Every MCP server you build starts from this TypeScript foundation. ## Agentic AI Deploy autonomous AI agents to ship internal tools and automations without writing code yourself. - [Coding with Agentic AI for Operators](https://www.fractionalskill.com/guides/agentic-ai-for-operators): Ship internal tools and automations without writing a single line of code yourself. ## AI Workflows Build repeatable AI-powered workflows that handle client work across your entire practice. - [Getting Started with AI-Powered Workflows](https://www.fractionalskill.com/guides/ai-powered-workflows): Transform AI from a chatbot into an autonomous work partner for your client portfolio. ## AI Skills Develop reusable AI skills and systems that deliver consistent results across every client engagement. - [Mastering AI Skills & Systems](https://www.fractionalskill.com/guides/mastering-ai-skills): Build reusable AI systems that deliver consistent results across every client engagement. ## Claude Agent SDK Build and deploy autonomous AI agents using the same engine that powers Claude Code. Covers tools, hooks, sub-agents, and production deployment. - [What Is the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-what-is-it): The same engine that powers Claude Code, accessible from code. Build personal assistant agents, business automations, and client-facing tools using the query function at the core of every Claude agent. - [Sub-Agents in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-subagents): How to delegate tasks to independent AI workers that run in parallel with their own context windows. Build planner-writer agent chains and multi-client research workflows that finish in a fraction of the time. - [Tool Hooks in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-tool-hooks): Intercept tool calls before and after they execute. Block sensitive file edits, build audit logs, and add guardrails that hold up in production without limiting what the agent can do for legitimate work. - [Built-in Tools in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-built-in-tools): The complete built-in tool set: file operations, bash, web search, sub-agents, task management, and plan mode. How to allow-list, block, and control which tools your agent can access. - [Lifecycle Hooks in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-lifecycle-hooks): Hook into session start, user prompt submit, and sub-agent stop events. Build a memory injection system that loads context automatically at the start of every session. - [Setting Up and Installing the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-setup-installation): Initialize a Bun project, install the SDK, and write your first working agent in under 10 minutes. Covers streaming input mode, the query loop, and what each option actually does. - [Building Custom Tools for the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-building-custom-tools): Create tools that call external APIs, define their input schemas with Zod, package them into in-process MCP servers, and give your agent capabilities no built-in tool provides. - [Permission Modes in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-permission-modes): Five modes that control how the agent handles tool calls: default, acceptEdits, bypassPermissions, plan, and dontAsk. When to use each and how tool lists interact with permission modes. - [Deploying a Claude Agent SDK Agent to Railway](https://www.fractionalskill.com/guides/claude-agent-sdk-deploy-to-railway): Package your Slack agent in Docker, configure persistent volumes so sessions survive redeploys, push to GitHub, and get your agent running 24/7 in the cloud with Railway. - [Project Setup and the Agent Chat Function](https://www.fractionalskill.com/guides/claude-agent-sdk-project-setup): Build the foundational agentChat function that every interface layer — Slack, terminal, API — calls the same way. Covers project structure, streaming input mode, and session ID handling. - [The Query Function: Two Modes Explained](https://www.fractionalskill.com/guides/claude-agent-sdk-query-function): Single message mode vs streaming input mode. Why the async generator unlocks hooks, session resumption, and multi-turn conversations that plain string prompts cannot. - [System Prompts in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-system-prompts): Three ways to write system prompts: plain string, the Claude Code preset, and the preset with a custom append. How the invisible base prompt works and what your prompt actually controls. - [What Are Hooks in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-what-are-hooks): Hooks run functions at specific points in agent execution. Understand tool hooks vs lifecycle hooks, how matchers work, and where hooks fit in the tool evaluation order. - [Multi-turn Conversations and the Interactive Agent](https://www.fractionalskill.com/guides/claude-agent-sdk-multi-turn-conversations): Build an interactive terminal agent with a conversation loop, session memory across turns, and a thinking indicator. The session ID pattern that makes follow-up questions work. - [Model Selection in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-model-selection): Haiku for development, Sonnet for production, Opus for maximum capability. How to set a fallback model and why switching models mid-session via hooks is possible. - [Session Management in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-session-management): Resume sessions, fork conversations from specific messages, and list session history. The session ID patterns that power multi-turn agents and persistent conversation tools. - [Settings Sources in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-settings-sources): How to load project CLAUDE.md files, skills, and settings.json hooks into your agent. The three source layers and why most SDK agents only need the project source. - [The canUseTool Callback](https://www.fractionalskill.com/guides/claude-agent-sdk-can-use-tool-callback): Programmatic control over every tool call. Allow, deny, modify inputs, and log tool usage with fine-grained precision that permission modes alone cannot provide. - [Structured Outputs in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-structured-outputs): Get typed, schema-validated objects back instead of text. Define a Zod schema, configure outputFormat, and build data extraction pipelines that feed directly into your workflows. - [Connecting the Claude Agent SDK to Slack](https://www.fractionalskill.com/guides/claude-agent-sdk-connecting-to-slack): Create a Slack app, configure bot scopes and socket mode, install Slack Bolt, and wire up channel mentions and DMs. The full integration from app creation to first response. - [Connecting External MCP Servers to the Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-external-mcp-servers): Give your agent a browser, a Notion workspace, or a GitHub integration by connecting external MCP servers. Faster than building custom tools for integrations that already exist. - [Polishing the Slack Experience](https://www.fractionalskill.com/guides/claude-agent-sdk-polishing-slack): Strip mention tags, convert standard markdown to Slack formatting, and add a thinking indicator. Three improvements that turn a working integration into one people actually want to use. - [Safety Limits: maxTurns and maxBudgetUsd](https://www.fractionalskill.com/guides/claude-agent-sdk-safety-limits): Cap how many turns an agent can run and how much it can spend per session. Two options that prevent infinite loops and runaway costs in any deployed agent. - [Message Types in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-message-types): The system, assistant, and result message types that every agent run produces. What each contains, when it fires, and how to use them for session tracking, cost logging, and error detection. - [Session Persistence with SQLite and Bun](https://www.fractionalskill.com/guides/claude-agent-sdk-session-persistence): Map Slack threads to Agent SDK session IDs using Bun's native SQLite. Build a session store with get, set, and delete methods that gives your Slack agent conversation memory. - [Skills in the Claude Agent SDK](https://www.fractionalskill.com/guides/claude-agent-sdk-skills): Give your agent reusable prompt modules it loads on demand. Build a meeting notes skill, enable it with the skill tool and settingSources, and create a library that scales across every agent you deploy. - [Production Hardening Your Slack Agent](https://www.fractionalskill.com/guides/claude-agent-sdk-production-hardening): Add thread context injection, thread locking for concurrent messages, and spending caps. The three steps that separate a demo deployment from one your team can rely on. - [Session Transcripts for Debugging Agent Workflows](https://www.fractionalskill.com/guides/claude-agent-sdk-session-transcripts): Find, read, and analyze the JSONL transcript files the SDK writes for every session. How to use a coding agent to parse transcripts and diagnose unexpected behavior in production agents. ## Optional - [Full markdown corpus](https://www.fractionalskill.com/llms-full.txt): every guide concatenated for one-shot ingestion. - [Sitemap](https://www.fractionalskill.com/sitemap.xml): machine-readable URL index. - [API catalog](https://www.fractionalskill.com/.well-known/api-catalog): RFC 9727 linkset describing public endpoints. - [Agent skills index](https://www.fractionalskill.com/.well-known/agent-skills/index.json): agent-facing skill docs. - [Newsletter signup skill](https://www.fractionalskill.com/.well-known/agent-skills/newsletter-signup.md): how agents should call POST /api/email. - [Site overview](https://www.fractionalskill.com/.well-known/agent-skills/site-overview.md): agent-friendly map of canonical URLs.