Alight. So lets say we have a semi retarded little kid. Let's call him Ash. Now lets say his big brother the hacker gives him a rating 6 comlink with the usual hacking suite and an agent that looks like some pokemon or whatever.

Is the communication between the kid and the agent such that he can hack as any other hacker would, except he'd be using the agents rating instead of his own non-existant hacking skill? The only disadvantage I can see to this is not being able to use edge. On the other hand you just got immunity to black IC's nastier effects.

Ditto technomancers I suppose.

I don't like it as it feels like a bypassing of the skillwire rules, and pretty much makes the hacker character a retarded option as any troll street sammi with logic 1 and zippo in computer skills can be just as good.


I'd kinda figured that a remote sprite or drone had to operate on it's own without much in the way of back communication. Limiting it's effectivness, since it would only be able to follow it's origional limited orders (dog brain style). But I don't think I could back that up.

well the agent would be loaded as part of the persona. as in, the kid is still "present" inside the system, but gets the agent to do all the work for him.

so, if he goes hot VR he still have to watch out for black IC and similar.

but under a setup like this he would probably be going AR anyways.

you may hit him with drone comprehension rules, as the kid may have trouble explaining to the agent what he wants it to do.

other then that, yes its fully doable.

Hacker in a box, cheap, easy, very useful.

Ah I think I just thought of something preventing rampant use.

Spoof.

Against a regular hacker you have to track them/dump them in cybercombat etc etc. Against an agent you could just spoof them in all the myriad amusing ways hackers spoof drones. And then some. Right?

*spoof* display the location of your controlling comlink via it's GPS system.

*roll roll*

Agent. "Ok!"

And since the user is relying on the agent for their matrix perception, commands like "do not display the icon of the spider in front of you" could also be fun.

spoof is one thing, the other is that if one go by the faq, every program the agent runs count towards the comlinks system rating. thats on top of the agent itself as it counts as a program on its own.

so if your running a agent + 5 programs inside the agent, your response drops by 1 if you have a system 6 comlink.

there are some "insane" maths floating around...

I'll double check when I get back to a book but I'm pretty sure that the roll for finding the traffic isn't any worse than the roll for the spoof.

Anyway the idea is that the kid wouldn't be present in the node that the Agent is in. The kid would be in his comlink and would simply be communicating with the agent. A bit like if he called up his hacker budy via a comcall and was telling them to do stuff.

To get at the kid you'd normally have to run a trace right?

Also for hacking on the fly I don't think the kid could follow the agent into the node, since hacking on the fly just gets through and does not provide a password or exploit. Or since the agent is running on the kids comlink does that also get him through if they wanted?

so in effect the agent is on a remote job for the kid?

ouch, there are so many ways that can go wrong

I think the persona rule applies if you're trying to spoof an agent when there isn't constant two way communication. At which point a rule more similar to what you do with a rigger would seem to be in order, maybe using tap comcall. If there isn't two way communication you're back to the thing operating remotly with it's dog brain, which limits what it can acheive.

(I'd be more specific but I'm away from my book)

As a seperate issue what nefarious things could you do to an agent you find wandering about? Obviously you can crash it. And if it's running off a comlink you could track it. But is there any way to hack/edit the thing? After all it may be running on a system you have admin privilages on.

Spoofing is pretty clear in RAW. You can't pretend to be someone without knowing who they are and thats where the matrix perception comes in. Otherwise it would be like trying to use the disguise skill without seeing a picture of the guy you were pretending to be.

I don't think there is a lot of things you can do to an agent that is wandering around on the matrix. I don't think there are rules in RAW for hacking or editing a program yet. Like everyone always says, maybe unwired will fix things. It would totaly be a house rule but you could go with something like if you crash the agent in cybercombat and it was loaded onto the node you were in you could edit it and reprogram it and then release it back into the matrix. Programing usually takes a long time so if it was a major change you wouldn't be able to do it but if it was to say write into the code to accept orders from you or to occasionally send all information it collects to random drop point or something like that then i'd go for it.

Intercept Traffic is what will get the kid in trouble, as his orders can get blocked or edited.

Ah yes exactly! I thought there were rules for it. That's what I meant by spoofing.

Since the Spider/IC at the target node have access to it they go straight to decrypt (if neccesary) and then the fun starts!

Though if it's just operating in dog brain mode I guess there isn't too much aside from it not properly handling getting traced or the like. And even then the node the thing is running in might not directly point to the kid. Though I guess if you traced it you could wait for the kid to contact to it or vice versa and intercept the traffic.

If the kid tries contacting the Agent which he will need to do in order for the Agent to actually take action then the nodes Hacker could potentially Edit the orders. ( Nothing like trying to tell your Agent that you need it to move to another node and the Hacker decides to Edit that to have the Agent go home .)

Another issue that can be a problem is other hackers ( Non-corp) that may decide to mess with the Agent or node the agent is in. ( Nothing like a little script kiddie getting miffed, trigger the node and then heading for the hills.)

Given how hacking is usually marginalised or game dominating, I really have no trouble with this. It means that a character can put a third of his resources into agents and a defensive commlink setup, and then spend the rest of his resources on being a street sam or drone rigger or even a mage.

The idea of a more "self-sufficient" type of sam who can mix it up via proxy on the Matrix is a kinda cool one. I'd have recommended it to my players if we didn't have a guy who's very fond of technomancers.

Yeah, but that just further marginalizes decking. You lose either way.

Well:

1) If the kid ever gives 'orders' to the agent (or is getting feed-back from the agent), then an intercept and edit means the agent now does what ever I want (possibly even 'deactivating' ('or better yet coping itself (moving/re-installing/etc) to my commlink).

2) if the kid never talks to his agent once released, then have the kid in english say what he wants it to do. Roll Agent Rating dice and the number of hits is the rough intelligence it has to processes those orders. Something simple like "hack this node and trash it' very likely will be followed. Break in and shut down the cameras without being detected is very likely going to get the agent confused and overwhelmed. (well shutting down the camera incrases the chance of getting detected, so I have to wait until nobody is monitoring the system then I can shut down the cameras safely, check good plan all parameter met).