← All Posts

How to Keep All Your AI Tools in Sync

· Lore

You start a conversation with Claude about your project architecture. Twenty minutes in, you've given it the full picture — the database schema, the auth decisions, why you chose that particular API pattern.

Then you switch to Cursor for implementation. It knows nothing. You start over.

Later, you open ChatGPT to draft a document. Same story. Same context. From scratch.

This is the default experience of working with AI in 2026. Every tool starts from zero. Your project's lore — the decisions, the research, the history — lives scattered across docs, conversations, and your head.

Every project has lore

Think about everything you know about your project right now:

  • The architecture decisions and why they were made
  • The user interview where Sarah said exports are broken
  • The meeting where the team chose Stripe over Square
  • The survey results from last month
  • The code patterns and the reasoning behind them

That's your project's lore. It's the accumulated knowledge that makes your project yours. And right now, none of your AI tools know any of it.

The fragmentation problem

Knowledge workers spend a significant chunk of their day just searching for information they already have. With AI tools, the problem is worse: you're not just searching, you're re-explaining.

Each AI tool maintains its own isolated context:

  • Claude knows what you told Claude
  • ChatGPT knows what you told ChatGPT
  • Cursor knows what's in your current file
  • Your coding agent starts fresh every session

There's no shared layer. No way for one tool to benefit from what you told another.

Collecting your project's lore

The fix isn't giving every AI tool access to every conversation you've ever had. That would be noisy, slow, and imprecise.

What you need is a single place where your project's important knowledge lives — searchable by meaning, accessible from any tool, and always citing the original source.

That's what Lore does.

How it works

1. Feed in everything

Drop files into a synced folder, push content from any AI tool, or pipe it through the CLI:

lore sync add --path ~/project/docs

Or just tell any connected AI tool:

"Save this architecture decision to Lore."

Meeting notes, user interviews, research, data exports, design docs — it all goes into one searchable knowledge base on Lore Cloud.

2. Search from anywhere

Every AI tool connected to Lore can search your project's lore by meaning — not just keywords:

"What did we decide about authentication?"

Lore finds the relevant documents across every source you've added, regardless of which tool originally stored them.

3. Get the full picture

Unlike AI memory that compresses everything into vague summaries, Lore returns the original text with citations. Your AI can tell you exactly what was said, where it came from, and when.

And with project briefs, you get a living summary of where each project stands — generated automatically from the totality of your lore.

Setting it up

Lore connects to any AI tool that supports MCP (Model Context Protocol) — which includes Claude, Cursor, Windsurf, VS Code, and many more.

Setup takes about 30 seconds:

npm install -g @getlore/cli
lore setup

Your knowledge base is stored in Lore Cloud and synced across every machine you use. The service is currently free — you just bring your own API keys for embeddings and research.

The result

Once Lore is running, every AI tool knows your project's lore:

  • Switch from Claude to Cursor? Your project history is already there.
  • Start a new coding session? Your agent picks up where you left off.
  • Need to write a product doc? Your AI cites the actual user interviews.
  • Want to know where things stand? The project brief has the full picture.

No more re-explaining. No more context lost between tools.

Ready to try it? Get started with Lore in under a minute.