Skip to content

Welcome to Knowledgeville

Knowledgeville is a public knowledge base: research on the things I find interesting, shared openly so it's useful beyond me.

Working with AI every day produces knowledge, but most of it stays in a silo: one person's, not shared. Knowledgeville tests whether it can work differently. Bundle it up, and you can plug it straight into your own workflow instead of rebuilding it from scratch. Connect your AI to it, and it has more to work with right away. My research becomes your starting point.

I work with Claude on all of it, as a standard workflow: take a question, research it, design the bundle, write it, publish it. It's anchored on the Open Knowledge Format (OKF) v0.1 as a test of the proposed standard. Right now that's frameworks, playbooks, assessments, and references, but I keep adding new ones, and expanding existing ones, as I explore new questions and ideas.

It's incomplete and rough, but it's been useful to me, and I think it will be useful to you too. Hopefully you'll help me test my theory, and let me know your thoughts.

Raff

Three ways to use this

Read it like a site — search for a concept, follow the tracks, use the assessments to benchmark where you are. Everything here is written for practitioners, not academics.

Feed it to your AI agent — the entire knowledge base is available as a public GitHub repository, structured in a format AI tools understand natively. Clone it, add it to a Claude Project, or point any AI assistant at it. You get curated, validated research as context rather than generic training data.

git clone https://github.com/grandaha/knowledgeville.git

Connect it over MCP — point your AI assistant straight at the live knowledge base. Add the Knowledgeville server as a custom connector. Then, from your chat, your assistant can search the concepts, open a page with its sources attached, and check what has changed. Always current, nothing to download.

https://www.onesteplabs.com/knowledgeville/mcp

In Claude, add it under Settings → Connectors → Add custom connector, paste the URL, and it is ready in any chat. Any client that speaks the Model Context Protocol can connect the same way.

Licensed CC BY 4.0 — free to use, share, and adapt with attribution. Cited third-party statistics and trademarks remain property of their respective owners.

Knowledge bundles

  • Open Knowledge Format — the open format this whole knowledge base is built on, explained for leaders: what OKF is, why a format for knowledge matters, and where the idea could go. Start here to understand how everything else is structured.

  • Enterprise AI Transformation — an eight-track model for taking AI from strategy to measured value across an enterprise: strategy & leadership, governance & risk, data readiness, platform, workflow, adoption, talent, and measurement.

  • AI Accountability — a role-by-role map of who owns which AI outcome inside an organization, with an operator's manual for each role, from the Chief AI Officer who sets the strategy down to the specialists who own risk, governance, and security.

  • Decision Guides — cited, plain-language guides for the recurring, consequential decisions a knowledge worker faces: the real options, the tradeoffs, the evidence behind them, and when each one wins.

  • AI Use Cases — practical, hands-on ways to put AI to work, grouped into sub-domains for different people and jobs. Most are copy-paste recipes — a ready-to-run prompt plus a before-and-after — for everyday tasks, executives, and realtors. The newest, AI Playbooks, go a step further: procedures an AI assistant runs for you across your email and calendar. More sub-domains over time.

  • AI Briefs — plain-language briefs of the AI work you won't read in full, in two kinds: research briefs of notable papers (what a study found and how much to trust it) and industry briefs of analyst and market commentary (an argument, and where the author's stake shapes it). Each covers what it says, how much to trust it, and what it means for you.

  • Team Second Brain — how a small team turns everyone's individual AI use into shared, compounding knowledge without buying an enterprise platform: what's already out there (The Single-Player Ceiling) and how to build the team layer yourself (Building Yours Without an Engineer).

  • AI-Assisted Software Development — the two real questions a developer or team lead faces with AI coding agents: does it actually make you faster, and does it open you up to new attacks. Covers the contested productivity evidence (Does AI Coding Actually Help?) and the named, documented vulnerabilities and campaigns already targeting AI coding assistants (Securing the AI Coding Workflow).

  • AI-Assisted Data Analysis — how to actually analyze data with AI, past pasting in a file and hoping: getting your data in without silent truncation (Getting Your Data In), catching the fabrication and arithmetic errors that survive real code execution (Doing the Analysis), and turning a result that worked once into a method you can trust again (Making It Repeatable).

Where to start

  • New here? Read What OKF Is — the open format everything here is built on, in a five-minute read.
  • Want something you can use today? Copy a ready-to-run recipe from Everyday Tasks, or the role-specific sets for Executive Leadership and Realtors.
  • Leading an AI effort? Read the Enterprise AI Transformation model — eight tracks that take AI from strategy to measured value across an enterprise.
  • Keeping up with the field? Skim the AI Briefs — plain-language reads of the research papers and analyst pieces that matter, with how much to trust each one.
  • Working with an AI assistant? Connect the whole knowledge base over MCP — see Three ways to use this above — and let it search and cite as you go.
  • Or use the search box (top of the page) to jump to any concept.