Help - Search - Members - Calendar
Full Version: Army of agents
Dumpshock Forums > Discussion > Shadowrun
kerbarian
Is there anything in the rules that prevents the following tactic?

A starting character buys the following software and cracks the copy protection: System 4, Agent 4, Firewall 6, Exploit 4, Attack 4, Armor 4 (total cost 27,000¥). He then buys 10 Meta Link commlinks and upgrades the Response of each to 4 (total cost 21,000). He sets up the agent software to run independently on each commlink (and provides copies of all the other software).

Now whenever the hacker wants to break into a system, he brings along his 10 buddies. The agents load Exploit, Attack, and Armor, and everyone hacks into the target system at once. Once inside, the agents attack any IC or security hackers present, and the hacker is then free to go about his business with 10 guards at his back.

Sure, the system will detect all this and be on alert, but is there anything it can do about it besides try to shut down? Tracking might be a problem, except that any IC/hacker trying to run a trace would probably have its icon destroyed before it got very far.

Similarly, why couldn't corps just do the same thing? Buy 1,000,000¥ worth of commlinks and load them with IC. Then, whenever any node on any of their networks has an alert triggered, they send in 100 instances of the IC to beat them down.

One potential limit is that when loading an agent on an independent node, "the agent doesn't count toward your persona's active program limits like running programs do, but it does count as a subscriber toward your subscription limit". That could be taken to mean that agents can't operate unless they're subscribed to a (metahuman?) persona. Even then, you'd have a ratio of 10 agents for every system-5 hacker. That limitation also doesn't make much sense to me, given that agents "will continue to operate in the Matrix even if your persona goes offline."

Finally, hackers probably wouldn't even need to buy the commlinks to run the agents on. Just go around hacking into random people's commlinks and installing your agent. Or even program the agents to hack in and install copies of themselves on any node they can find. It would be like modern botnets.
Konsaki
It wouldnt work because any GM worth his salt would smack said player up side the head for being a twinking retard. If you want a true answer though, do a search for Serbitar's matrix PDF. This question has also been answered before, and I'm not inclined to try and answer it agian.
kerbarian
QUOTE (Konsaki)
It wouldnt work because any GM worth his salt would smack said player up side the head for being a twinking retard. If you want a true answer though, do a search for Serbitar's matrix PDF. This question has also been answered before, and I'm not inclined to try and answer it agian.

I realize it can be prevented by GM fiat, but I was hoping for an answer that would be more understandable to the players. Sure, it's powergaming, but it also seems like a perfectly realistic thing for a hacker to try. I'll read through that pdf, though -- Thanks.
dog_xinu
my TM in my group is doing something similar but different. He has had the group subscribe all their commlinks into a group-wan (gwan from here on). Now he has lots of drones (only uses a few at a time unless he puts them on autopilot). He feeds all his information from the drones to the other players via the gwan. When he does matrix searches and such and get info back, he relays it to the group via the gwan. the one thing he has done and he wasnt 10% clear with the group about before he started doing it, was using everyone's commlinks as storage devices. So when he is hacking something he defers the "file" to various people's commlinks to expand his storage. One of the team didnt understand what he was doing and took it as an attack. That was just a misunderstanding that the characters (and players) worked out.


So he is not doing the 10xCommlinks == 10xAgents to do stuff. But he is doing the gwan thing. He is also cracking all knowsofts that they can find and spreading them out to everyone. You know, spreading the knowledge.


dog
Jaid
QUOTE (kerbarian)
Is there anything in the rules that prevents the following tactic?

A starting character buys the following software and cracks the copy protection: System 4, Agent 4, Firewall 6, Exploit 4, Attack 4, Armor 4 (total cost 27,000¥). He then buys 10 Meta Link commlinks and upgrades the Response of each to 4 (total cost 21,000). He sets up the agent software to run independently on each commlink (and provides copies of all the other software).

Now whenever the hacker wants to break into a system, he brings along his 10 buddies. The agents load Exploit, Attack, and Armor, and everyone hacks into the target system at once. Once inside, the agents attack any IC or security hackers present, and the hacker is then free to go about his business with 10 guards at his back.

Sure, the system will detect all this and be on alert, but is there anything it can do about it besides try to shut down? Tracking might be a problem, except that any IC/hacker trying to run a trace would probably have its icon destroyed before it got very far.

Similarly, why couldn't corps just do the same thing? Buy 1,000,000¥ worth of commlinks and load them with IC. Then, whenever any node on any of their networks has an alert triggered, they send in 100 instances of the IC to beat them down.

One potential limit is that when loading an agent on an independent node, "the agent doesn't count toward your persona's active program limits like running programs do, but it does count as a subscriber toward your subscription limit". That could be taken to mean that agents can't operate unless they're subscribed to a (metahuman?) persona. Even then, you'd have a ratio of 10 agents for every system-5 hacker. That limitation also doesn't make much sense to me, given that agents "will continue to operate in the Matrix even if your persona goes offline."

Finally, hackers probably wouldn't even need to buy the commlinks to run the agents on. Just go around hacking into random people's commlinks and installing your agent. Or even program the agents to hack in and install copies of themselves on any node they can find. It would be like modern botnets.

sure. those agents are totally, completely useless to you.

1) they're going to get you detected on the way in, with only 8 exploit dice, unless you're hacking a crappy, pathetic target anyways.

2) those agents don't have analyze or stealth. it's not a question whether they will be noticed or not. they will be noticed. they will set off the alarm. and with that many agents in the node, the node *will* shut down. i don't care what you do, there's still a hardware switch around. additionally, the only person who has any chance whatsoever of spotting IC, security deckers, etc is *you* none of your agents can do so. any security deckers are no more at risk of being noticed than they would have been had you shown up. in the event all agents switch from exploit to analyse, i would use the rules for group perception; you get +1 die per additional person spotting (assuming you have the best dice pool, that is), to a maximum of +5. useful, yes... useful enough to justify the fact the server will be shut down via hardware by the next combat round because it is blatantly obvious it's under a huge attack? hell no. enjoy your dumpshock

3) your agents could all be spoofed by a skilled decker.

now admittedly, these get a lot less worrisome if you go with rating 6 agents... but then you're looking at a rather large investment, and i would still probably only give you the teamwork benefit except for engaging in cybercombat. incidentally, you were doing it all wrong btw... subscribing to each agent separately means you must give them each a separate command to do something. subscribe to them as a group if you just plan on ordering them all to do the same thing instead, and you save subscription slots.

on to the rest: sure you could have your agents install themselves on someone else's commlinks. but stop and think: who has rating 4 (or, if you want *useful* help, instead of captain sucky's worthless pieces of crap, rating 6) commlinks? not an awful lot of people... and most of those people will be able to do something about a rating 6 agent showing up on their commlink. things like tracing back who is responsible, and loading hostile agents onto your commlink, to beat you into a bloody pulp as soon as you show up. possibly your own agent, because they will be able to remove their wireless and take their time dealing with the agent. even if it's just some rich guy who arbitrarily wants a high rating system, do you really think he's not gonna spend some money defending it?
Charon
QUOTE (kerbarian @ Dec 25 2006, 09:45 AM)
Finally, hackers probably wouldn't even need to buy the commlinks to run the agents on.  Just go around hacking into random people's commlinks and installing your agent.  Or even program the agents to hack in and install copies of themselves on any node they can find.  It would be like modern botnets.


Well, to this at least I can reply that it's unlikely to work well because most people have crappy commlink that can't run a decent agent (Operatings system too low).

And if you start hacking into tough commlink in order to make them run your agents so that can use them in tough network... It's not very productive.

---

It's also silly. What's the point of barging in like Attila the Hun and his horde? The system adminstrator will opt to shut down over any other options as soon as he sees your legion and then what have you accomplished?

---

There are no rules for it, but personnally if a guy is running tons of program that are a copy of each other, I'd diminish their efficiency and would attack them as a group somehow.

In terms of programming, they are all identical line of codes that all try to exploit the same weakness in the target and all have the same weakness to be exploited in themselves. That's good enough fluff justification to impose various penalties anyway.

It also can't be that hard to track them. None of these agent is especially good at matrix perception so unless you spot the the trace I don't see how it's less likely to work than in any other scenario. Though I might be wrong, Matrix ain't my specialty.
Serbitar
Lots of agents = very bad idea

Hacking is about stealth. Once you are detected, the system can simply shut down and the run is over. You do not want to be detected. Lots of agents hacking in WILL be detected -> run is over.

Furthermore I add the following houserule:

Every matrix perception test concerning a persona in the same pass yields a -2 dice pool modifer (cumulative).

Same goes for matrix attack.

Explanation: You can smuggle only so much bad code into a hackers data trail at one moment.
kerbarian
QUOTE (Serbitar)
Lots of agents = very bad idea

Hacking is about stealth. Once you are detected, the system can simply shut down and the run is over.

Several people seem to be saying this, and I agree. However:

If this tactic works very effectively to shut a system down, that in itself is a quite powerful tool for runners in some situations.

It also still doesn't address why corps wouldn't use it for defense. If it only costs a few thousand nuyen for a system that can run decent IC, why wouldn't the megas have thousands of such systems so that they can dispatch at least dozens of IC agents to deal with every system alert? The IC systems could all be running in a central location, and the agents just log into the system with an alert, wherever it is.

I suppose you could say that they do have thousands of such systems, but there are also thousands of breakin attempts triggering alerts at any given moment, so the IC is still spread pretty thin. That's not really the impression I got from the fluff, though.
hobgoblin
ah, good old ddos. i wonder what unwired will have to say about that one nyahnyah.gif
ElFenrir
QUOTE
Lots of agents = very bad idea


While after reading varied hacking rules, i can see why...question is, I know agents have their good uses. From anyone who has played an SR4 hacker...what in your opinion is a 'safe' number of agents? (I know either way, rating should be high, of course.) A couple of agents to help out perhaps?

I can think of above million-agent scenario as the ol' 'too many cooks spoil the pot' saying somehow.
Serbitar
The number of agents that does not increase the possibility of being detected too much is good.
Jaid
also, maybe it's just me, but... wouldn't an agent not on the commlink you're using be stuck to their own commlink? iirc, unless it's with your persona, it can't go places really. it's pretty much stuck where it's sitting, i thought...
Serbitar
thought wrong
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Dumpshock Forums © 2001-2012