You protect your identity much better by taking care not to leave ID traces than by resleeving - but that requires information hygiene, which is hard for most characters and players. I am currently running an agent-campaign, and my players have some downright hilarious and smart methods for contacting each other, using dead drops for items, and generally staying hidden in false identities. Both players and characters are pretty good at "tradecraft". Yet certain enemies do track them because of small, apparently innocuous mistakes done way back. The main reason these enemies have not attacked is that the time was not right or they had other things to do - it is a big and messy world. Now, the above agent game is pretty hands-on. The PCs are simultaneously field agents and assets, largely because their agency has kind of broken down. If I were kind enough to give them peace and quiet, they would no doubt set up layers of assets to do the dirty work. The problem game-wise is of course that the dirty work is the really fun part, so beyond one or two missions where things go as they should, it is more rewarding to have everybody running - their agency compromised, their allies untrustworthy, their ultimate loyalties still untested. Much more tension, drama and surprise. In another EP campaign the PCs just got a wonderful cover: a fairly powerful player gave them untraceability as a boon for giving it a xrisk-level piece of information. They are protected from tracking to some extent, although it is not perfect and they should maybe have scrutinized the exact wording of the promise. This was introduced by me partially to deal with the fallout from a previous mission where logically various agencies would start looking for them - even hyperelietes (nay, *especially* hyperelites) should not be involved in luxury habitats blowing up. My most stable EP campaigns actually have people who either do their job (the game is about a particular job like gatecrashing or salvage) or have a solid little organisation that makes sense (philanthropic project to help Tanzanian infugees). In this case a lot of "fun stuff" can be hidden inside the facade of the organisation and its activities, with some PCs acting as outward faces and the more dynamic ones as internal troubleshooters.

I don't think Firewall would use murderhobos. I certainly wouldn't, if I had their resources and motivations. What I would do is to build up a large Ego database. These can be purchased through shell companies from criminal organizations or from unscrupulous Hypercorps, or acquired by various forms of Egonapping. Then, whenever I needed any covert work done, I'd select one or more appropriate Egos from the pool, and groom them for the mission. This would include high-end psychosurgery, accelerated Simulspace training- and indoctrination sessions, pruning and repeated mission simulations, using all data acquired from the best surveillance and data gathering methods available. The key is to leave nothing to chance. The agents must have exactly those competences, backgrounds, personalities, ideologies, behavioural tendencies and limitations, pieces of (mis)information and orders which make them the most likely to achieve successful mission outcomes. Leaving anything to chance - or worse yet, the whims and strange logic of murderhobos - is very likely to result in someone using these same methods against you to figure out who you are and oppose you in some way. I can see two ways to get PCs in on the above fun. One would be to play the support for Firewall agents. You need to render favours on short notice, while otherwise minimizing your involvement with and knowledge of the situation. You might or might not know the name Firewall, or you might know the organization by another name, or think you're part of some different organization entirely. You'll occasionally see something really creepy or improbable, particularly in terms of what Agents look like when you're in close contact with them or when they're returning from a mission, or maybe as you're called upon to clean up an operation site. And of course, if the mission goes completely to hell (as even the best projections might credibly prove inaccurate, when TITAN tech is involved,) then you might end up in danger. The main focus of such a game might be your everyday life in a Transhuman universe, punctuated by occasional encounters with and assignments issued from Agents of Firewall, which range from the merely odd and improbable to the horrifying and life-threatening. The other possibility is of course that the Players play out the operation in a wider sense. You present them with the mission to be undertaken and whatever intel they have available, and allow them to plan it out, and create characters specifically for the purpose of playing that mission. They then take on the roles of these characters, and play them as they carry out the mission, whereafter they may freely replace or modify those characters after they've had the next mission briefing. If the players all agree that the number of Egos required for the job is different than the number of players, then you might co-opt something like the Troupe dynamics from Ars Magica, to allow players to share characters or control extra characters collectively. This isn't really as munch

One thing to note about Firewall is that to a large degree, the concept doesn't lay itself too solidly to the idea of the common roleplaying games. A Firewall sentinel seems like someone who would be activated in a manner that is conducive to their normal living, such that the subtlety of the conspiracy can be maintained. To that end, Firewall would likely pick candidates whose skillset and connections can actually aid the organization, without any significant need for said sentinel to change their normal habits and routine. Of course this isn't conducive to traditional party-based roleplay (unless all characters are close companions in-setting, maybe), so the more common playstyle I've seen tables turn to is something like Shadowrun, where sentinels are professional criminals working for Firewall. Which, let's be honest, means you're a potential murderer that travels from place to place as Firewall needs without a home to necessarily go back to... a different take on the "murderhobo" concept. Now this doesn't have to devolve into a cliche, mind you. A group like Firewall will inevitably need crimes and atrocities committed (it does sort of go with the "prevent another Fall at [b]all costs[/b]" mindset), so it makes sense that it would have operatives for that task. To that end, I don't see why a sentinel group couldn't make money and gear a second priority to the mission at hand... Firewall needs resources, after all, and anything a sentinel has is resources for a Firewall asset. But with how ephemeral physical goods tend to be handled in-setting, being money-hungry isn't a very useful option. What good is a nice apartment and the slickest morph on the planet if you're egocasting a few AU around the system for most of the year? As for whether such agents can maintain anonymity at all times, the likely answer is "hell no". A sufficient panopticon will be all but unaware, and characters will leave some trail at some point. But to some degree, this is what the Vectors are there for. To maintain broader information security on Firewall missions. When that's not an option, Firewall operatives have to instead come off as something that the system and public are less likely to care about. For example, in one of our games the group had to mask their activities as politically-subversive crimes by radicals. When someone is always watching, your best option is to simply make what you're doing seem expectable.