I imagine that if you wanted to make one from mostly "off the shelf" parts you'd basically need a well stocked Solarchive (blueprints, science, and human history, etc). Getting good blueprints is critical, for the seed to thrive I imagine that you'd want good blueprints which include alternate feedstocks in case of material shortages, and a lot of designs. Power generation, survival gear and habs for any number of environments (a lot can change in 10,000+ years), as well as everything else needed to bootstrap up transhuman civilization, a task too large to really elaborate here. A desktop nanofab would be my choice for fabrication, a general hive has a massive failure point in that if the conditions or feedstocks necessary for a Prothean hive aren't right, the entire operation fails. A well hardened cornucopia machine (with a really long duration power supply, maybe a burning lens, chemical oxygen generator, and some kind of stable fuel like crude oil, powering a fuel cell?) might be best. A CM can make more power generators, more of itself, and anything else with blueprints). It'd be terrible if EM interference or strong winds or something stopped the protheans. A well hardened small server which can hold the entire population as static ego files (easy) and run a few (harder), which will still work (call it a Specialized Server) should handle the population. Summing that up, you'd be looking at: * General Solarchive [High], I made that price up, but a specific one is [Low] A normal Solarchive is about smartphone sized size [small], but this might be more like a tablet. * Blueprints [Extremely Variable] I don't want to make a price for this, because it really varies by how much basic infrastructure is available from orgs like the Argonauts. Call it 100k credits to be safe and get good and redundant blueprints. You'd only need to spend that once though, as they can be used repeatedly. * Long Life Desktop or Industrial CM [Expensive (30,000+)]. These are normally cheaper, but I figure one which can be relied upon 10,000+ years with maintenance or power is a lot more expensive. It'd probably require a lot just to get it working again after that much time, I imagine the nanoscale parts are fragile, even to stuff as small as thermal fluctuations and noise. Normally about a cubic meter in size, but probably bigger with the extra shielding and chemical power supply. Call it 2 cubic meters, or .056 cubic meters for a hardened general hive. * Enhanced Server (Long life) [50,000 credits for 10 infomorphs]. Using the portable server as a guide, and doubling it for .08 m^3. Costs total to 185,000 credits, assuming you need 100k worth of blueprints, with a second unit cost of 85,000 credits once the blueprints are bought. Or about 120k credits needed for the blueprints for the hardware. These are pretty spitballed prices, but they are based on the prices in the books, so they're probably ok to use. For the big design you're looking at ~2.1 cubic meters of volume, and .136 m^3 for the hive design. (about a bathtub, or larger personal luggage).

It'd be pretty easy to upgrade a Kudzu pod with a strong enough rocket to go to other planets in the system should be pretty easy, especially on gates on planets with a low escape velocity. Going to other solar systems would probably be a lot harder, but obviously not impossible, because that's exactly the kind of stuff being done at Justin Case. The rocket wouldn't need to be as overengineered as the rest, as it doesn't need to last as long. Stealth in space is very hard, and essentially impossible with longer journeys. Intermittent engine burns can help, but there's a lot of eyes watching in the solar system, and those engine burns will heat the ship itself to visibility. Looking like a piece of debris, or a ship which is supposed to be there helps a lot. There's a big difference between someone seeing a small rocket and someone caring though, so long as you stay away from important places.

Mass drivers heat their projectiles. Eclipse Phase's near perfect metamaterials probably offer some interesting ways to direct waste heat, but I guess that they aren't likely to be good enough to hide a ship completely.