Project Brain
A second brain for any project. Docs and code.
Point it at a repo and it reads the docs and the code, then builds a wiki you can ask questions of. How does checkout reach the payment service? Why Postgres locks instead of Redis? What breaks if you touch billing? Ask the brain instead of grepping for an afternoon.
It is one skill folder that drops into any project and runs inside Claude Code. It only ever writes to a .brain/ folder, so it is safe to run on a client codebase. Little Brain organizes your whole operation; Project Brain understands one project at a time, and you run a separate brain for each repo.
Three steps. No drama.
Init
It scans the repo, proposes how to file things, and scaffolds a .brain/ folder. You confirm and you are done.
Ingest
It reads your docs and your code and turns them into a linked wiki. Run it again whenever things change.
Query
Ask a question in plain language and get an answer with links to the pages it came from. This is where you live.
What it does.
- 01Reads docs and code: models modules, decisions, and gotchas, not just files
- 02Ask-the-brain queries that cite the exact pages they came from
- 03Drops into any repo. One skill folder, config-driven, no vault required.
- 04Write-isolated to a .brain/ folder, so it is safe to run on a client codebase
- 05Self-serve at $99, or have us set it up and run it, from $1,500
The full kit.
- 01The skill itself: init, ingest, query, and lint, all config-driven
- 02Code ingest that models a codebase into module, decision, and gotcha pages
- 03A how-it-works guide and a best-use-cases doc, so you can run the whole method yourself
- 04The two-tier offer: a complete self-serve product, with hands-on setup available when you want it
Project Brain is the portable version of the second brain that runs a working multi-brand studio every day. The method is proven before you download it.
The honest answers.
How is this different from Little Brain? +
Little Brain is a whole-life vault you fork once to run your entire operation. Project Brain attaches to a single project and reads its docs and code to build understanding of that one thing. You run Little Brain once. You install a Project Brain for every repo you want to understand.
Do I need Claude Code? +
Yes. Project Brain is a skill that runs inside Claude Code. It needs the project it is reading and nothing else: no server, no account, no database.
Is it safe to run on a client codebase? +
Yes. It can only write to a .brain/ folder it adds next to your code. It never edits your source, your config, or your git history.
What is the setup tier? +
Done-for-you. We stand the brain up on your actual codebase, tune it to your system, run the first ingest, and can stay on to keep it current. The self-serve docs are the same ones we follow; you are paying not to have to.
Project Brain.
In the final stretch of the build. Pre-order locks your copy for release day.