Projects and knowledge
A workspace where you keep the files, instructions and chats for one body of work together, so every conversation in it starts with the same background.
Also called: Projects · Spaces · notebooks · knowledge base · project files · workspaces
Connections
Understand first
Ideas this one builds on.
Leads on to
Ideas that build on this one.
Related
Often used together or easily confused.
Blueprints that practise it
Resources that teach it
- Claude 101 (garden)
- Claude Cowork 101: How to automate your workday without touching code (How I AI, with JJ Englert) (garden)
- The Agent That Saved My Brain (garden)
- AI Style Guides: How to Help AI Write Like You (garden)
- How to Use Claude Code as a Second Brain (garden)
- Compound Writing: The Ultimate Guide (garden)
- Build your personal AI copilot (garden)
- Lenny's Data: free public starter pack (garden)
- How to build your PM second brain with ChatGPT (garden)
- Behind the Craft — Transcripts (garden)
Competency
Find projects and knowledge on the commons map ↗
You may know it from
- law a case file Everything about one matter in one folder, so anyone picking it up starts from the same facts.
- office work a shared project folder
In products
The same idea, under different names in different tools.
| Provider | What they call it |
|---|---|
| Anthropic | Claude Projects |
| OpenAI | ChatGPT Projects |
| Gemini Gems with files; NotebookLM | |
| Perplexity | Perplexity Spaces |
In plain words
A project is a folder for AI work. You put in the documents that matter (a style guide, a strategy, past reports), write a few instructions for that work, and then every chat you start inside it already knows that background. You stop re-uploading the same files.
The products have different names. Claude and ChatGPT call them Projects, Perplexity calls them Spaces, and Google spreads the idea across Gems with attached files and NotebookLM notebooks. Underneath, they all do the same job: keep context for one area of work in one place.
Choose the files with care. A small set of current, relevant files works better than a dump of everything. Out-of-date documents get treated as true. And because files in a shared project may be visible to teammates, think about what belongs there.
Example
A grants manager creates a project for one funder. It holds the funder’s guidelines, last year’s application, the budget template and the note “always match their language about outcomes.” Each new section she drafts starts from the same facts.
How it connects
- A practical tool for context engineering, combined with project-level custom instructions.
- Close cousin of memory: projects keep chosen files, memory keeps learned details.
- Custom assistants package a similar setup for others to use.
Try it
List the three files you paste into AI most often and guess which one will make the biggest difference. Put all three in a project and try one of your usual requests.
Read this card as plain Markdown: /concepts/con-projects-and-knowledge.md. Know a parallel from your own field, or a product that implements this? Add it.