Shadowrun
Shadowrun General => Gear => Topic started by: Dal Thrax on <06-24-15/1534:46>
-
You can slave rating * 3 devices to a comlink, RCC or cyberdeck. You can order devices from your PAN. A RCC makes it easier to control drones, but from my reading of the rules anyone can control a drone (using its internal pilot program) from a simple comlink. By my reading, you can still give orders to devices slaved to your PAN so.
Fairlight Caliban ($8,000) + 4 Nixdorf Sekretar ($16,000). The Caliban runs a browser app while the agent from each Sekretar performs an aid another action on a matrix information test. Result + 6 dice - 1/2 time.
Fairlight Caliban ($8,000+ 21 Nixdorf Sekretar ($84,000). The same idea but instead you run 7 clusters of 3 agents. The first agent in each cluster makes a gather information search, The other two use aid another. This means each cluster is running Agent Attribute (3) + Agent Skill (3) + Aid Another Bonus (3 [on average]) or 9 dice on a gather information search.
Cyberdeck with Level 6 agent + 3 Nixdorf Sekretar's per agent. In this case each agent program runs Agent Attribute 6 + Agent Skill 6 + Aid Another bonus 6 (on average) dice to gather information.
Attack Dog (Clearly Rules Legal), Nixdof Sekretar + Level 2 Attack Dongle ($16,000). Probably best used to aid another in cybercombat.
Cyberdeck + 10 Attack dogs ($160,000). A swarm of agents isn't exactly subtle, but gets the job done.
Sleaze Dog (Clearly Rules Legal), Nexdorf Sekretar + Level 2 Sleaze Dongle ($16,000). Probably not all that useful for the don't get seen aspects of sleaze, but handy for cracking files and hack on the fly.
Upgraded Dog (might not be rules legal). Transys Avalon + Program Module + Level 6 Agent ($17,900). Assuming an average of 4 successes you would want a level 4 dongle ($65,900 total).
Upgraded Firewall. Fairlight Caliban + Program Module + level 6 agent ($30,900). A fairlight Caliban running a agent with orders to perform matrix overwatch (matrix perception tests) and perform the full matrix defense action if the device is attacked. 12 dice (Rating 6 attribute + rating 6 willpower) + applicable device rating (most likely firewall) or 19 dice to resist most matrix attacks without the character needing to spend an action.
Plus B&E ($78,900), the same as Upgraded firewall but with a level 4 sleaze dongle giving 10 dice to resist matrix perception when running silently.
Extra Cheese, any of the above plus a machine sprite running diagnostics on the central device.
Possible rules issues: The ruling that a device persona merges with your person when in VR might cause the device chain to drop. I would think that you would still be able to set up an agent swarm around a comlink then access it through the control device action.
Fun extra: You know agents are so good at one of the primary tasks of computing (finding information) that corporate bigwigs are going to want to run with whatever allows them to control the most agents. Could be some interesting flavor if the characters break into a stock brokerage and everyone is running either high end cyberdecks or RCCs to max out on the number of agents they can control.
Suggested house rule: limited the max number of agents that a character can control to either their Computer or Leadership scores.
-
You can only order one autopilot at a time with each simple action. The device where your autopilot is located on does not have to be on a device that is slaved to a master device. It just have to run on a device you are the owner of. This can be done when you form your persona on a commlink.
If you form your persona on a RCC then you can instruct multiple autopilots on different devices, but in this case the devices need to be slaved to the RCC you based your persona on.
Not clear if you need to order them all at once to benefit from a teamwork test (in that case you are limited to use a RCC as a master device, if not then you can get an unlimited amount of autopilots). Also not clear if they can assist you in a teamwork test at all.
-
I'm not sure about drones (though if I was GMing it I would say yes and tell the rigger that the drone fires off three rounds when aiding in combat, though I could also see a drone sensor to smartgun link). With agents, you're giving order that amount to follow me (or follow him) and help with any X task that he undertakes. Seems valid. Especially if the the orders are to follow somebody and help with finding information on the matrix. That's sort of the intended function of commercial agent programs.
One interesting question would be the ability of a rigger to modify some form of attack dog to work as part of the drone network. Physical drones, check. Matrix drones, check. Cost issues probably help keep things, somewhat, under control, but I have to wonder if a couple comlinks with agents might be close enough to a decker. Also, it allows a technomaner to run both agents and sprites, which might get a bit out of control.
-
One (slight) limiting factor on Agent nets making teamwork tests is that if any of them get a critical glitch, you don't get any bonus to the main character/agent's (deck's?) limit.
-
Also, dice rolling hell. Be prepared to duck the GM swinging the rulebook at you...
-
Um, dice rolling app?
-
In that case only resources stop you. Buy 100 commlinks. Modify them all with a program carrier containing an agent. Buy 100 sleaze dongles. Spend 100 simple actions telling them to assist you when you hack the host. Hack the host.
They don't have electronic warfare, so I don't think you can tell them to snoop you and just assist whatever matrix action you take (multiboxing)
-
arent teamwork tests limited by the most dice that can be added to the test is equal to the leader’s rating in the applicable skill, or the highest attribute rating if the test involves two attributes. so max 6 dice bonus course if the mage casts analyze device and or technomancer diagnostic
-
You're right on that Rooks.
The most dice that can be added to the test is equal to the leader’s rating in the applicable skill, or the highest attribute rating if the test involves two attributes.
-
So you can still have a 100 assistants, just that you will -at most- only double your effective skill rating (and the risk for glitch is quite high).
Probably more reasonable to have ~6 higher rating agents or ~12 lower rating agents (running on separate commlinks; where half of them have an attack dongle and the rest a stealth dongle).
No need to slave any of them.
-
Could you give the agents a contingent order like "I raise my arm, everyone hack the host and look for this file" then just throw a 100 agents at the thing? I guess that gets us into questions about IC. Does the Host get one IC per rating for each intrusion scene (so a bunch of deckers can overwhelm a hosts IC) or one IC per rating per intruder (which would weaken agents and sprites), one IC per rating per intruder that is not a sprite (makes technomancers more powerful but leaves the question of what happens if you just send in a sprite) or one IC per rating per intruding device (which would seem to take care of many of these problems, but would require some careful phrasing as registered spirtes are independent entities, meaning that you could end up nerf technos even further into the ground)?
Even if hosts can fight back with IC, there's still the issue of setting up an army of information gathering agents. It's kind of hard to justify nerfing something for doing its designed function.
-
The benefit of this is that you can have agents erasing marks running automated interferance while you can deal with other more pressing issues course the downside is the overwatch score and in typical SR5th fashion no clue if agents have their own Overwatch Score or if its shared by the device (my ruling is overwatch is like monitoring an IP address so its all under one score)
-
Well sense the agents are running on their own commlinks, I'm pretty sure they have individual overwatch scores. Since agents on your deck share all of its traits, I'd say that one running locally probably shares your score (and can add to it).
-
Course if you really want to get creative get a RCC slave all the commlinks to the RCC then send out one command to all/select few commlinks to do various actions
-
Can you use an RCC to control agent programs? If so, you're going to get some interesting results, for example stock analysts on Wall Street using RCCs to command a hoard of research agents.
-
Well Pilot programs are just like agents
-
Well Pilot programs are just like agents
Well, very similar, at the least. Pilot programs are designed to pilot something, and Agents to work the 'trix. The differences, from a software stance are academic, and their mechanics are quite similar. Still, running multiple Agents on an RCC would run into the same problem as running them on a 'deck; cannot run to identical programs. So I would think Agents would have to have different ratings, thus further diminishing the overall usefulness of the tactic.
Just my .02¢
-
He is talking about running agents on multiple cyberdecks (or customized commlinks) and then slave
them to your RCC and use the unique feature of the RCC to send the same command to multiple autopilots as one simple action to command each group of agents with one simple action each.
-
Well Pilot programs are just like agents
Well, very similar, at the least. Pilot programs are designed to pilot something, and Agents to work the 'trix. The differences, from a software stance are academic, and their mechanics are quite similar. Still, running multiple Agents on an RCC would run into the same problem as running them on a 'deck; cannot run to identical programs. So I would think Agents would have to have different ratings, thus further diminishing the overall usefulness of the tactic.
Just my .02¢
actually Praetus you can run identical programs on the RCC, to a point.... how else would you get 2 maneuvering autosofts for 2 different types of drones running at the same time? granted their not identical because you need to have them made for the drone they are supposed to control, (roto-drone, steel lynx, dalmation, etc) similar to how a r1 agent is identical to a r6 agent in every aspect aside from die pools.
the question remains is that difference in die pools enough to make the argument that they are different programs? I'm unsure as to that.
and yes i know im basically doing a circle on myself with this comment...... but meh