Arsenal Kills Agent Smith, Ding Dong the Exploit is Dead... Options

[Ryu]

i just have to ask, have anyone seen someone attempt to deploy agent smith tactics in game?

Not in action because it could be averted, but I have seen the calculations for an array of comlinks (compareable harmless as this Agent Smith was only supposed to defend the drone network).

[Cthulhudreams]

I've never had anyone try to deploy it because its been explictly disallowed since day 1.

But lots of people acting entirely reasonably (including me!) do the tacking an agent on to an otherwise hack free character concept.

[mfb]

i just have to ask, have anyone seen someone attempt to deploy agent smith tactics in game?

in SR3, yes (having played only half a game of SR4 somewhat handicaps my ability to relate SR4 anecdotes). we've got a rules lawyer who realized exactly how powerful he'd be if he just created one badass agent and uploaded multiple copies of it everywhere he went.

we also have a long-played rigger/decker character who employs what can only be considered small armies of drones, managed through careful use of CC drones which come with their own RC decks. using the custom drone rules, he's even created infiltrator drones which can detect, decrypt, and subvert enemy drone networks on their own. it would be ludicrous if he weren't exceptionally subtle with their use (unlike the Agent Smith decker above).

[Dashifen]

I've never seen anyone try to pull an Agent Smith in SR4.

[ixombie]

in SR3, yes (having played only half a game of SR4 somewhat handicaps my ability to relate SR4 anecdotes). we've got a rules lawyer who realized exactly how powerful he'd be if he just created one badass agent and uploaded multiple copies of it everywhere he went.

we also have a long-played rigger/decker character who employs what can only be considered small armies of drones, managed through careful use of CC drones which come with their own RC decks. using the custom drone rules, he's even created infiltrator drones which can detect, decrypt, and subvert enemy drone networks on their own. it would be ludicrous if he weren't exceptionally subtle with their use (unlike the Agent Smith decker above).

Agents were actually more expolitable in SR3, because you did not have subscriptions. You could upload as many copies of an agent wherever you wanted. But in SR4, it clearly states that each agent you have acting independently of you counts against your subscriber limit. Why? Who knows. But it sure helps to prevent one hacker from taking over the whole matrix...

The best way to deal with players who do completely legal things like use armies of drones or (at least in SR3) armies of agents is to punish them with the consequences hammer. The more allies you use, the more likely the enemy is to capture one and use it to find you. There's no such thing as the perfect, untraceable hack. If there were, hackers clever enough to write worms which infect the entire world would not be going to jail all the time.

[mfb]

i'm not a big fan of punishing players for things which make sense both IC and OOC. Agent Smith makes sense in both categories--as i said before, the Storm botnet is a perfect example of a real-life application of the Agent Smith strategy.

agents count against your subscriber limit in SR4, just as registered drones do in SR3. in SR3, it was possible (or at least believable and not specifically disallowed) to equip a drone with an RC deck and then subscribe the drone to your own RC deck. you command the drone, the drone commands other drones. i'm not aware of a specific reason you couldn't pull the same trick in SR4. if there is a specific ruling that disallows this, it's stupid--there's no sensible reason why it shouldn't be possible in the game world. yes, it breaks the game--which, to me, means that the game rules need to be altered to allow it in such a way that it doesn't break the game, rather than simply disallowing it.

hardcore rules enthusiasts will argue that the game's reality, not the rules, should be changed. that's okay for them, but i honestly can't imagine an internally-consistent reality that allows drones to work they do in SR4 without also allowing Agent Smith.

[kigmatzomat]

I think the "decker in a box" aspect of Agents could be eliminatedif there was a penalty for having anything other than common use programs loaded. Reduce their rating by 1 for every hacking program installed, with the associated limit conveyed down to the programs.

Ta da, they could go back to being dog-smart AI useful for people who aren't so good in the net. Bob in accounting can use his agent for finding the TPS reports, wiz-mage Steve's agent runs the apartment automation but HackMaster Ralph finds that his agents are just about useless when loaded with Exploit, Stealth, and Spoof. Sure, it was an R:6 agent before but now it's R:3, along with all the apps.

Of course, I thought the existing rules were fine. You can run a hacker-agent on your Comm but that pretty much means you aren't doing anything. Or you can have an agent logged into the target machine but you need to exploit/probe an account for each agent. There's a lot of chance to trigger an alert just from them logging in.

edit: I may be off on an assumption. I'd thought that if you and an agent were both in the comm then the total program count was user's programs + agent's programs + 1 (for the agent) with the spoof/stealth/firewall/etc being shared since they are tied to the single connection. So typical hacker defenses (spoof & stealth) + agent + agent's active program(s) + hacker's active program(s) came to a minimum of 4 programs, more likely 5, and 6+ is real probable.

I'm not sure that's the case after re-reading the rules. Nobody's really played a decker in SR4 and I've been trying to avoid memorizing the rules since Unwired will change stuff around.

[mfb]

eh, that would help some, but there is still a hell of a lot that an agent can do for you with purely legal programs. to use an SR3 example, if you give one of your agents a superuser account, it can give a valid superuser account to every other agent you upload without even rolling.

[masterofm]

@ hobgoblin - Well I don't know I think even if you try to shut down a drone in the SR setting (because hackers can be a problem) is that even if you do bring down a node for a little while, generally that node will transmit a distress call that basically transmits "AndroidXV804 Log 18209: Hacker w/ multiple high lvl agents. Raise threat lvl to condition red. Rebooting to alternate mode." Maybe you bring that node down for a little while, but when it restarts it is a closed system and now everyone in the area will be gunning for you, because wireless hacking (unless a player seriously backs up his game [and most don't]) will be found out pretty quickly. If they are at a node or a terminal then they have just seriously compromised their position too. Yeah there are a few cases where using the agent smith tactic to shut a system down is effective, but why not let them get away with it if it only works in a few select cases? People crash servers with bots. I remember back in the very beginning of WoW some people were angry at how bad warriors were so they created tons of gnome and dwarf bots and crashed my server.

Also I would like to just point out Agent Smith is not the win, I mean Neo killed him in the end.

[Moon-Hawk]

Here's how I run things. None of this is canon, but I haven't had any issues with Agent Smith:
First off, multiple copies of an agent don't contribute anything to a task, they're completely redundant. Multiple copies can be sent in different directions to do different things, but if attempting to cooperate or even accomplish the same task in parallel, they are 100% redundant. Roll once, that's how they do.
Second, I assume that hosts don't suffer the standard response degradation for having users connect. It just doesn't make any sense to me.
Third, I assume agents can access host A while being resident on device B (just like a hacker) or load themselves completely onto host A, as described.
Fourth, remember that Agents have to be loaded with programs, so if they've loaded themselves onto some foreign host (as opposed to a program-laden commlink) they've got a limited set of programs to work with.
Fifth, I have, in the past, limited Agents to rating 4. Since Augmentation, I've removed that limit and put the max back at 6.
Sixth, the fact that a high-rated Agent is better than a piss-poor hacker is a feature, not a bug. I'm betting that any serious Hacker will be throwing more than 12 dice, especially considering the toys in Augmentation. (bear in mind I use Logic+Skill for hacking) I don't want any player to be forced into playing a hacker in order to have a successful game. I like that the team can buy an agent to do some light hacking for them.

None of it is official, and I'm not sure it makes Agent Smith impossible, I'm just saying I haven't had any problem with it. But I realize that I use lots of matrix house-rules, so my experience with them has virtually nothing to do with how they really work in the game.

I'm hoping Unwired will sort everything out. A lot of people say that the matrix is so horribly broken that they can never possibly be fixed without contradicting SR4, which we know they will not do. In response, I say look at what they did in Augmentation with cyberlimbs (without contradicting the main book, mind you), and have hope.

[ixombie]

i'm not a big fan of punishing players for things which make sense both IC and OOC. Agent Smith makes sense in both categories--as i said before, the Storm botnet is a perfect example of a real-life application of the Agent Smith strategy.

Well, punish isn't the right word. More like balance. Just like if someone launches a massive DDoS attack using a bot network, launching a huge attack with lots of agents sets of lots of alarms. Private and government net security people will be all over it like a swarm of bees. If you just quietly hack something and take what you want, there's not going to be some huge alarm. But if you use the matrix equivalent of a fullscale invasion or nuclear bomb, people will notice, and people will come after you. It's the same thing with drones - if you attacked someplace with 50 drones, the target as well as law enforcement will try their best to find out who you are and punish you. But if you just have a few drones on an in-and-out job, it won't be their top priority.

The idea is that when players do something that is legal, yet over the top, the game world reacts accordingly. Nothing says you can't have an army of drones or agents. But also nothing says that these kinds of hamfisted tactics are smart, or that you will get away with using them. Becase they aren't, and a good GM won't let you.

[Spike]

@Spike

Or, you know, as its an RPG, and the players are supposed to be part of a team, with different abilities to combine to form narrative threads, and I can tell the mage player to go stick it, and ban agents, just as I'd tell the hacker to get stuffed if he wanted an ally spirit, without subjecting himself to all the rigmarole normally required to summon one.

Also your analogy fails as the typical security drone (I imagine that to be a prime target too for mage hackers with agents) can only support 8 dice of counter hacking due to the system & response 4 of a steel lynx. Boned.

The r6/r4 chip thing being talked about before also doesn't work because plenty of hacking attacks can be launched without being loaded on. Exploit springs to mind. So does spoofing as the registered owner and sending the shut down command

I don't buy it. When I play a game, I generally am the front line combatant. I don't feel my toes are stepped on when the Mage whips out a gun and shoots a mother-f**ker in the face. I DO get upset if the Hacker tells me I have to fight all twenty sec cops because 'not my job, man'.

The hacker in a box isn't going to replace a dedicated hacker on the team, but if the GM is setting up runs knowing the team has additional hacker support (like... what if two players REALLY wanted to play hackers? Which one do you make suck it up and not play one so the other guy gets to keep his niche? Same problem, really...) then there are plenty of 'lesser matrix tasks' that the Mage can run his HIB through while the dedicated hacker handles the 'really hard shit'.

As for the 'limited response' issue, there are a couple of points. As I recall from the last round of debates on the Matrix I participated in, it is open to interpretation wether or not the agent is restricted to the ability of the node or not. I thought the common concensus was yes, even if it doesn't make perfect sense, but I think your counter example proves you fall into the other camp.

Not that it matters. The mage is still not doing magic shit when he tells his HIB to go hack that drone over there. Sure, the agent is using ITS IPs to actually DO the hack, and the mage had better hope that the Drone isn't being actively rigged.

And provided the GM wasn't stupid enough to set up the run where hacking that Steel Lynx was meant to be the Hacker's 'BIG MOMENT' of the night, then the Hacker probably will be happier he has one less thing to worry about while he is busy trying to prevent the arcology blast doors from closing and locking the party in, or what-ever.

[mfb]

Well, punish isn't the right word. More like balance. Just like if someone launches a massive DDoS attack using a bot network, launching a huge attack with lots of agents sets of lots of alarms.

only if the agents are crashing around setting off alarms--and even then, it's only a maybe. the decker in question actually made an impressively stealthy agent and just uploaded it a million times (SR3 rules). their presence didn't trip any more alarms than the decker himself did.

and like i said, it's only a maybe even if they aren't ninja-sneaky. the Storm botnet (i love this damn thing), for instance, has set off alarms all across the globe. but there are so damn many instances of the thing, only a tiny fraction of which poke their heads up at any given time, that the best effort to eradicate it only managed to reduce its projected numbers by 20%.

[Cthulhudreams]

A few things.

As for the 'limited response' issue, there are a couple of points. As I recall from the last round of debates on the Matrix I participated in, it is open to interpretation wether or not the agent is restricted to the ability of the node or not. I thought the common concensus was yes, even if it doesn't make perfect sense, but I think your counter example proves you fall into the other camp.

err, what? if the agent is loaded onto the steel lynx's node of course it is affected by the steel lynx's response. just that at no point does the action spoof actually require it to be loaded on the steel lynx's node. Decrypting a data bombed file would.

Not that it matters. The mage is still not doing magic shit when he tells his HIB to go hack that drone over there. Sure, the agent is using ITS IPs to actually DO the hack, and the mage had better hope that the Drone isn't being actively rigged.

And provided the GM wasn't stupid enough to set up the run where hacking that Steel Lynx was meant to be the Hacker's 'BIG MOMENT' of the night, then the Hacker probably will be happier he has one less thing to worry about while he is busy trying to prevent the arcology blast doors from closing and locking the party in, or what-ever.

actually I doubt the mage would be wasting his time issuing orders, when you can just tell the agent to spoof a fake order to any security drone that starts shooting at me to shut down, and then give the agent a live feed from the mages googles so it can see whats shooting at you. At R6 It is considerably smarter than a steel lynx drone, which is more than capable of understanding complex concepts like that, but asteel lynx can understand orders like 'atrol the complex, and alert us if you find any intruders.

[Lyonheart]

I've decided my solution to this problem is simply to strictly enforce the hackers ability to command agents. That should keep it relatively in control, besides I like to throw around a lot of IC.

As for the Hacker in a box, every member of the team needs one, otherwise the Mage, Street Sam, etc. are going to be constantly tring to keep my agents out of there Comlinks, so I'm cool with that.

In fact I'm fairly positive when I get my game going the other PC's are going to put there Comlinks at the disposal of the hacker, so that doesn't take away his job, those hacker in a boxes all run on what is effectively his home network.

[Cthulhudreams]

I've decided my solution to this problem is simply to strictly enforce the hackers ability to command agents. That should keep it relatively in control, besides I like to throw around a lot of IC.

As for the Hacker in a box, every member of the team needs one, otherwise the Mage, Street Sam, etc. are going to be constantly tring to keep my agents out of there Comlinks, so I'm cool with that.

In fact I'm fairly positive when I get my game going the other PC's are going to put there Comlinks at the disposal of the hacker, so that doesn't take away his job, those hacker in a boxes all run on what is effectively his home network.

isn't the best defense against this sort of stuff to just have a 1/1/1/1 commlink that you don't care about, and turn it off during runs, and disable the wireless on everything else you have, using a skinlink, googles and mic to control it?

[ixombie]

only if the agents are crashing around setting off alarms--and even then, it's only a maybe. the decker in question actually made an impressively stealthy agent and just uploaded it a million times (SR3 rules). their presence didn't trip any more alarms than the decker himself did.

and like i said, it's only a maybe even if they aren't ninja-sneaky. the Storm botnet (i love this damn thing), for instance, has set off alarms all across the globe. but there are so damn many instances of the thing, only a tiny fraction of which poke their heads up at any given time, that the best effort to eradicate it only managed to reduce its projected numbers by 20%.

No agent is trace-proof. And a lone hacker trying to pull a big global attack does not have anywhere near the resources that THE ENTIRE WORLD has. They will find you, and they will lock you away and throw the key in a plasma furnace. Shadowrun doesn't take place in 2008. Corporations and governments have been working full-steam to quash cybercrime ever since Echo Mirage, since they have experienced (TWICE) how one super virus can crash the whole net.

If you accept the premise that anyone with an agent program and some knowhow can spread their agents everywhere across the whole matrix, one of two realities follow from that. Either a) there are agents everywhere, poised to crash the whole matrix, or surgically destroy any given system at any time or b) the authorities know how to deal with that kind of crap, and people who try it are usually thwarted and thrown in jail where they become the village bicycle for trog prison gangs.

Which one is more supported by fluff? I sure don't recall anything about how the world is held hostage by botnets, and that hackers rely on swarms of agents to do their work. Everything I know suggests that hackers rely on their own skills and their own hardware and their own 1337ness. There must be a good reason for this. We can infer from the state of Matrix fluff that people who try Agent Smith do not get away with it. The fact that people can get away with it in 2008 has absoultely no bearing on whether they get away with it in fictional 2070.

[mfb]

actually, any agent can be trace-proof, just as any decker can be trace-proof (or near to it). all you have to do is, before you hit your target, jump around to different RTGs around the world--low-rated ones, where you can do your thing without fear of detection. perform a Redirect Datatrail operation on each. voila, you have now increased the TN of any Track or Trace attempt made against you by up to +8 (depending on how many RTGs you did this on).

besides, even if they trace your agent, all they have is the host the agent was uploaded to. that could be any host in the world that you hacked into. there's no way, in SR3 (or SR4, that i'm aware of) to put a Track or Trace on someone once they've logged off or even just jacked out.

[Tiger Eyes]

actually, any agent can be trace-proof, just as any decker can be trace-proof (or near to it). all you have to do is, before you hit your target, jump around to different RTGs around the world--low-rated ones, where you can do your thing without fear of detection. perform a Redirect Datatrail operation on each. voila, you have now increased the TN of any Track or Trace attempt made against you by up to +8 (depending on how many RTGs you did this on).

besides, even if they trace your agent, all they have is the host the agent was uploaded to. that could be any host in the world that you hacked into. there's no way, in SR3 (or SR4, that i'm aware of) to put a Track or Trace on someone once they've logged off or even just jacked out.

In SR3, you could be trace-proof (I loved that!!!). SR4 is much more harsh. It's a simple Extended Computer+Track test, threshold 10, interval=1 initiative pass, to "follow the target's current connections node-by-node back to the source." (SR4, pg 219) If you're wireless, they'll be able to triangulate your position within 50 meters. And if you want to Redirect Trace, you must win an opposed Hacking+Spoof vs. Computer + Track/System+Track, with each net hit adding 1 to the trace threshold. And, even worse, you "can only redirect a trace in progress." (SR4, pg 224). Any Track programs that go after your agent, regardless of if they're in the same node as you or not, will trace back to your point of origin. (pg 227)

So, in SR4, you're no longer trace-proof. In fact, you're pretty much screwed as soon as an IC or hacker starts tracing you. Even worse if they start tracing your agent, since how would you ever know??? Until the friendly Lone Star boys are knocking on your door... (IMG:style_emoticons/default/grinbig.gif)

[Kyoto Kid]

..and this is why I say that the "shadows" have been taken out of Shadowrun. (IMG:style_emoticons/default/ninja.gif)

[Tiger Eyes]

..and this is why I say that the "shadows" have been taken out of Shadowrun. (IMG:style_emoticons/default/ninja.gif)

Oh, really, KK. Such melodrama. (IMG:style_emoticons/default/wink.gif) It really means that my hacker can finally wirelessly hack from a crowded starbucks in the middle of a crowded mall, sipping her latte while trashing systems, and knowing that there's five hundred other people with commlinks within 50 meters, and that Corp Security is gonna have a huge headache questioning all those shoppers... Or better yet, shop while hacking! Yes! Finally, multitasking at its best! (Out of the basement and straight to the mall, that's my theory (IMG:style_emoticons/default/love.gif) )

[hobgoblin]

one thing about tracing a agent, it only points to you if your running within your persona (or in other words, on your comlink).

and isnt one of the basics of agent smith that you upload agents to other nodes and run them form there, subscribed to you?

btw, as i reread the bit about agents, im getting the feel that the idea of uploading isnt that the agent jumps from node to node, but that its running on a node separate from your own comlink. it can do the jumping as well, its even talked about how it affects its response rating, but it does not need to (unless the node its on is going down asap and it need to continue working, making me think if big agent supporting clusters).

[Lyonheart]

isn't the best defense against this sort of stuff to just have a 1/1/1/1 commlink that you don't care about, and turn it off during runs, and disable the wireless on everything else you have, using a skinlink, googles and mic to control it?

Well, yes, but then you can't get any +4 bonus for AR. You can't see what your buddy's are doing with there eyes, I don't recall there being stats for encrypted frequency hoping radios that are not Comlinks or that would be anymore hacker proof.

I'd recommend them having a cheep through away Comlink as well, not for there runs, but for talking to there Fixers, Johnsons, Etc.

[Kyoto Kid]

Oh, really, KK. Such melodrama. (IMG:style_emoticons/default/wink.gif) It really means that my hacker can finally wirelessly hack from a crowded starbucks in the middle of a crowded mall, sipping her latte while trashing systems, and knowing that there's five hundred other people with commlinks within 50 meters, and that Corp Security is gonna have a huge headache questioning all those shoppers... Or better yet, shop while hacking! Yes! Finally, multitasking at its best! (Out of the basement and straight to the mall, that's my theory (IMG:style_emoticons/default/love.gif) )

...ahhh but where's the sense of adventure in that...? (IMG:style_emoticons/default/wink.gif)

I enjoyed having to sneak into a facility with my chummers, find a jackpoint such as a Sec Cam or automated SoyCaf maker, and then dodging IC though the Corp's matrix until I hit paydirt. Decks were unique, expensive & complex yes but still cool as they had style that commlinks lack. Also, I couldn't be hacked unless someone knew what system I was on found me and then back-doored into my deck while I was jacked in. It was also all about anonymity. My purchases (and movements) were harder to trace, I only really needed a fake ID when I had to bluff my way into some place while on a job. Shoot, I could go into an AAA neighbourhood as long as I dressed the style walked the walk, talked the talk and looked like I had the cred. (IMG:style_emoticons/default/smokin.gif)

Didn't have to know computers if I wasn't into decking, and never had to worry about some pesky RFID or Commlink pinging who and where I was everywhere I went. (IMG:style_emoticons/default/wavey.gif)

...yeah, part of me will always long for the "good old Neo Anarchist days". (IMG:style_emoticons/default/frown.gif)

[ixombie]

actually, any agent can be trace-proof, just as any decker can be trace-proof (or near to it). all you have to do is, before you hit your target, jump around to different RTGs around the world--low-rated ones, where you can do your thing without fear of detection. perform a Redirect Datatrail operation on each. voila, you have now increased the TN of any Track or Trace attempt made against you by up to +8 (depending on how many RTGs you did this on).

besides, even if they trace your agent, all they have is the host the agent was uploaded to. that could be any host in the world that you hacked into. there's no way, in SR3 (or SR4, that i'm aware of) to put a Track or Trace on someone once they've logged off or even just jacked out.

As Tiger Eyes points out, the rules don't support you on that, at least in SR4. But more to the point, think about the consequences of what you're saying. If hackers can make agent armies that can take over the world at will with impunity, then they would do it. The fluff doesn't support a new hacker taking over the world every day. Not by a long shot. No matter how you slice it, there is SOMETHING preventing massive botnets from controlling the matrix. The devs neglected to tell us what it is. But they also didn't tell us that botnets are the norm, that they are indeed the core of hacking. As the books explain it, the Shadowrun matrix is about the hacker going mano a mano with his target node. That's the matrix in the 2070s. It's not like modern day. It might not even be realistic. But that's how things work. The most logical explanation is that cyber police will kill people who try to create huge botnets, which deters anyone from even trying. But use whatever explanation you want. Unless you want to go Trollman style and recreate the Shadowrun matrix in your own image in terms of how you want it to be, you're wrong.