Context

Everything on this shelf is real — real engineering, real incidents, real fixes. But it's also a curated slice of a much larger, private wiki, and some things are left out on purpose, not by oversight.

What's deliberately missing

• The real network map. No addresses, no machine identifiers, no exact topology. A reader gets the shape of the design (why three zones, why one service is public and the rest isn't) without the specific map that would let someone actually target this network.
• Authentication and secrets, in any specific detail. Which credential does what, how a token is scoped, where a password lives — described in principle, never in a way that would help someone probe a real system.
• Anything about my employer. My day job sits in a regulated industry, and I've been careful, throughout this shelf, to describe what that means for how I practice security here without ever describing anything about the employer itself — no name, no system, no real data.
• The compliance and hardening documentation tied to that professional context. A separate, internal book exists for that material — reproduced homelab exercises inspired by real-world regulatory practice — and it stays internal, permanently, with no exception. It was never a candidate for this shelf, and it never will be.
• A large amount of routine, unglamorous internal documentation. Step-by-step runbooks, exact configuration values, service account inventories — accurate and useful to me, but not interesting or safe to show a visitor, and not rewritten for this shelf.

Why publish an editorial choice instead of just staying quiet about it

I'd rather say plainly what's left out, and why, than let a visitor wonder whether an omission is an oversight. A shelf that's honest about its own edges is more trustworthy than one that quietly hopes nobody asks. If you're a peer evaluating whether this homelab's practices are real rather than staged, the fact that some things are visibly, deliberately withheld is itself part of the evidence.

What you can trust is real

The incidents, the pitfalls, the numbers where they appear (durations, counts, version gaps) are the genuine ones, not smoothed over for presentation. Where something was rewritten for this shelf, it was rewritten to remove identifying detail — never to make a failure look smaller or a fix look cleaner than it actually was.