I picked up a copy of Mongoose's Agent sourcebook recently. My motivation in this was, in part, that at some point I want to run a conversion of Dark Heresy to Traveller (this project tentatively named Dark Travesty), and such a game would feature agent-type operations prominently. Overall I think this sourcebook has provided several features which would add to a game of this sort.
Agent has a fairly standard structure for a Mongoose class book. We get 25 pages covering six new careers, 10 pages on agencies, 17 on legal proceedings, 15 on espionage, 12 on corporate operations, 13 on bounty hunting, and 23 of new equipment (including deckplans for four ships and Bond-esque prototype equipment rules). The careers chapter is surprisingly good for Mongoose; usually I dislike these sections, as the careers are frequently either too specific to be useful or they overlap with each other. Agent has Law Enforcement, Investigator (a private-eye sort), Spy, Analyst, Corporate, and Bounty Hunter careers - effectively, it splits the Law Enforcement specialty of the core agent into Law Enforcement and Investigator careers, splits Intelligence into Spy and Analyst, leaves Corporate alone, and adds Bounty Hunter. Honestly I like this structure; it's nicely genre-appropriate for a washed up police officer to to become a private investigator, which is impossible under the core rules. Six careers feels like about the right number, too. The specialties within Agent's careers seem sufficiently distinct for the most part, and you can usually tell in the career artwork which character is from which specialty, though there are some exceptions. I am, however, displeased with the fact that the survival and advancement difficulties are higher across the board than those in the core book; while this is standard for Mongoose supplement careers, it rubs me the wrong way that a corporate technical expert agent who spends his time in a lab under these rules should have as hard a time surviving as a spy on foreign soil under the core rules. If I use these careers, I will likely decrease these difficulties, much as I do for the Scoundrel careers. Yes, crime and agency are both dangerous, but the Core establishes an expected difficulty for them which seems out of line with that seen here. While many of the Agent careers do have higher cash mustering-out awards than their Core equivalents, there are also those like Law Enforcement which use essentially the same table, and which still have their survival difficulties increased, so that justification does not cover the whole issue.
The other interesting point from character creation is the addition of Trust, Networks, and Cover Identities as awards during character creation (sometimes in place of skills). These provide benefits when using some of the mechanics detailed later in Agent, but overall I think they're somewhat underutilized, or perhaps more accurately that there are few to no details on how to use them outside of the game structures which much of the rest of the book consists of. Of these three mechanics, I think Trust is handled best. It rarely takes the place of a skill, accumulating mostly as a benefit of even-numbered ranks and on the Personal Development tables. When it appears on mustering-out benefit tables, it mostly appears in addition to another benefit (for example, Investigator benefit 1 is "+1 Trust, Contact"). Finally, the mechanics for using Trust are fairly well-developed in the Agencies chapter, for purposes as general as "Requisition 10,000 credits worth of equipment from the Agency." Thus, it has applications in most any situation a Traveller might find himself in (there are caveats to actually using these applications of Trust, but for those qualified, it looks quite nice). Networks and Cover Identities, on the other hand, do not appear to be linked to anything other than the subsystems detailed in later chapters, most prominently bounty hunting. There's no mention of (for example) substituting a cover identity in place of Intelligence for a Deception roll when acting under that cover, nor of using the strength of a network for research purposes except when bounty hunting. These sort of omissions are saddening. I will, however, almost certainly be stealing Trust for Dark Travesty; it seems a decent way to model the sort of influence which the minions of an Inquisitor might carry.
The next chapter details four sample agencies plus freelancers. Each of these describe what agents do within these agencies, and provides ways agents can utilize their Trust scores to gain benefits from the agency they work for. The flavor notes are mostly pretty good, but I'm saddened by the fact that the Trust functions are only usable by those characters who ended character creation as active agents. This may be realistic, but it does not sound exactly 'fun'. What if you leave an agent career to assume a Noble / Administrator position in the intelligence community, or some other extenuating circumstance? Fortunately, this is nothing GM fiat and being recalled to active service cannot fix.
Chapter three covers an extended version of the legal process outlined in the Law Level section of the Core, with a focus on players as the prosecution rather than as the defense. It opens with an elaboration of the law levels and categories to which they apply, which is not particularly helpful except for the clarification on Tech restrictions (LL9+ and restricted tech is described as "state-imposed medieval culture", which does nicely explain what TL restrictions mean practically). These are followed by the extended legal process, which has three phases focused on gathering evidence through the application of various skills, followed by a trial and expanded sentencing tables. After this comes a brief section on Informants, a variant on Contacts which provide very specific mechanical benefits. These sound quite strong for what they cost you - having a Market Assistant informant could net you a +1 DM to three broker rolls within 24 hours of each other at a cost of 50 credits, which will almost certainly pay for itself if used for the buying or selling of goods. A few artistically-rendered criminal dossiers round out this chapter; these are good seeds, but could have been put into a single page rather than the three that they span, and the art for them is somewhat lacking. Judgement: filler. Overall a decent chapter for those on legal offense, though.
Chapter four is espionage. This chapter feels somewhat chaotic, as the ordering is quite bad. It begins with a very simple espionage system, where a spy rolls several skill checks and totals his effect, which is compared with the effect of several defensive / countering skill checks made by opposing agents. OK, that's usable enough. Data analysis is discussed next; this, as it turns out, is useful to generate modifiers to espionage mission payouts as part of the espionage mission system at the end of this chapter. Between data analysis and the mission system, though, we have a discussion of elite 'hacker' agents, Imperial agents, the roles of various sorts of agents in espionage operations, and rules for what happens to agents when there's a regime change in their agency. This ordering seems distinctly non-optimal; data analysis should be in with the espionage mission rules. I found the hacking rules very objectionable (what does "Destroy opposing encryption with one-way entry" mean, anyways? And since when does developing any sort of workable encryption scheme take 1d6 hours? Deploying an existing one, maybe, but creating secure cryptosystems takes years of work by specialist cryptographers... and number theory was not on the list of skills Agent requires for hackerdom); I will be using Scoundrel's computer rules instead, I think. These hacking rules are not well-integrated with the other systems in Agent, in any case. I am a huge fan of the Imperial Licenses, though - these are pretty much exactly what I'm looking for for Inquisitorial purposes. I could do without the requirements for terms in particular careers, but other than that these look very usable, and the qualification process seems almost sufficiently stringent (add "must serve as an interrogator for a senior inquisitor until deemed ready" and it works for me). The discussion of the roles of various agents, while somewhat interesting, could have been put at the beginning of the careers chapter, since it does a good job of clarifying the differences between Investigators and Law Enforcement agents, among others. It is also not terribly specific to espionage. Likewise, Regime Change is not specific to espionage operations in particular, though it is interesting and well-developed. I would have put this in the Agencies chapter, since it deals primarily with the Trust mechanics and regards change of agency power structures rather than the performance of espionage operations.
Finally, we come to rules for just those operations. Much like mercenary tickets, these are presented as a way to fill down-time, and they have a structure which recurs in both the Corporate and Bounty Hunting missions of later chapters - roll assignment, which tells you which skills you will need to roll. Roll target, which provides modifiers to pay and difficulty. Roll mission duration, which determines how long it will take. Roll difficulty, which determines the total effect you need to score with the skills designated by the assignment. If your total effect is greater than or equal to that required, you succeed and roll for pay, and if it is less, you have failed and roll for a mishap on a table of unfortunate consequences. This "roll multiple skills and total your effect" notion is repeated several times in this book, and it does provide an interesting alternative to the task chain as a means of resolving long, complex actions. It is also very amenable to parallelization across a party along the lines of 4e's skill challenges, but using effect rather than boolean successes or failures. Just as with 4e's skill challenges, though, I'm not sure the math works here. Any espionage op is going to require checks with four specific skills. Across all seven types of op, 15 distinct skills are represented, and since the type is chosen randomly for each mission, it is quite possible that an agent will just not have one or more of the required skills (Trade? Seriously?), which is going to dump a big ol' -3 on one of his checks (yes, you do have to attempt all four...). Further, on average a mission requires about 8 effect to successfully complete. With average rolls of 7s on the skill checks, this means you need to be at +3 combined skill and attribute DMs for each skill to succeed on average, and it could be much worse; a difficult mission against a hard target might require 15 effect instead, which will require either spectacular rolls or about a +5 modifier to each of the four rolls to succeed. Between the semi-random assortment of skill checks required and the degree of competence required on average, I'm just not sure that it's possible for most agents to succeed with any degree of reliability. The payouts aren't particularly good either, averaging about 5000 credits for 2-3 weeks of work (though the payoff on corporate and bounty hunting ops are significantly higher), and failure tends to result in loss of Trust or attribute points. All in all, it seems a Hard Mode system, much like smuggling was in Scoundrel. If I were to use this, I think the agency would probably select missions based on the strengths of their operatives, rather than at random, since this would result in fewer casualties and greater rates of success, in which the agency has a vested interest.
The Corporate chapter is somewhat better-structured than espionage. It begins with a discussion of how corporate warfare is conducted, what ends it serves, and what tools are useful for it. Highlights of this section are stats and wages for corporate security forces and mercenaries. This is followed by a brief-and-general discussion of industrial sabotage, and then rules for corporations manipulating governments; I'm curious if these mesh with Dynasty or Merchant Prince. Some options for would-be infiltrators are considered, and then we tangent back out to the digital realm of electronic infiltration; again, I find their rules dubious. I don't have any good numbers on how long it takes to set up a good corporate spearphishing attack, but I strongly doubt that 35 weeks is the average. A list of security systems (mostly sensors) are given, along with lists of skills which could be used to bypass them; these are used to determine the skills rolled for Corporate missions, the rules for which follow immediately. The structure is very similar to that of Espionage missions, except that the list of skills you need to roll is determined by the security measures in place, you get a random number of rolls to make your effect total (on average you need 12 effect out of 6 rolls, so you still need +3 skill+attribute on average), and the payoff's a little better. Fortunately, the list of skills you really need is Stealth and Sensors; those two together will let you circumvent everything on the security systems list. It's much more viable to get +3 in each of those two skills than in each of the 15 skills you might need for espionage, especially since Sensors is wafer-jackable. In short, corporate work is a lot more viable financially than espionage. As icing on the cake, the pay is better and the missions are shorter, too.
Bounty Hunting is a dense chapter. It begins with general advice for would-be hunters, and then moves to necessary mechanic qualifications for a professional one. These are naturally strongly biased towards PCs created using Agent's rules; it would be fairly difficult for a Core PC to qualify, as they would have to be a 3-term Agent with Streetwise 2. All other avenues would be closed to them. A discussion of types of contracts follows - there is a bug here, as accepting a Private Contract nominally permits one to roll on an alternate column for bounty pay; no such column is to be found on the pay table some ten pages later. Rules for tracking ships across jump come next; these could be quite handy both for PCs chasing other people, and for determining how long it takes bounty hunters to track down one's PCs. Again, we encounter "Make several checks with an effect totalling X", but here there are many options and the real barrier is time, since there's a clock ticking down to see when your target will move again. Thus, you can gamble with increasing the time your checks take to get bonuses to your effect. Once you've tracked your target to a particular planet, you then have to locate the individual, and an analogous system for tracking people planetside follows, along with profiles for various sorts of likely bounty targets ranging from runaway noble scions to axe-wielding stim-heads. All of the rules so far in this chapter have been generally usable outside a dedicated mission structure, and are pretty good; tracking people down is useful for one's own ends too. Now we turn to assassination and bounty missions, though, where we're looking at 13 distinct skills where you need to average +3 in order to be able to succeed reliably. The payout is leagues better than espionage work for comparable difficulty, though - on average, somewhere around 30,000 credits. Honestly, this whole system is useful to me mostly for figuring about how much a bounty should be worth; I'm not all that interested in resolving bounties in this fashion, but knowing how much of a price my PCs have on their heads is always entertaining.
Finally, equipment. The prototype rules look pretty reasonable; if I ever get around to running Fantasy Traveller, this would not be a bad way to generate magic items. The prototype value table looks pretty iffy, though. Some sneaky weaponry and TL12 improved silencers follow, but there's some weirdness with the "needletto", which in its rules text has the note "The Effect of this weapon is always considered to be one point higher for the purposes of damage." I'm really not at all clear how this is different from adding 1 to the damage. If it were a laser weapon and the +1 applied to generating eye hits on effect 6+, that would be one thing. But it isn't, and this note just leaves me somewhat confused. Armored bodygloves make an appearance; these are par for Ravenor and Eisenhorn's retinues, and so I'm happy to see them here. There's a fair bit of weirdness and technobabble though the rest of the gear in this chapter, but I do like the presence of the Doctor Tam Special of drugs that make you seem dead (though there's no nausea side effect post-awakening; might have to add that). Testimonial Drones make a nice servo-skull replacement, but overall I think I'd need to do a fairly careful case-by-case consideration of which of these gadgets I want in my game. The book closes with ships. Frankly, I just don't get all that excited about ships and deckplans, and these were not exceptions. The fact that two of the four referenced other supplements (I guess Advanced Probe Drones were in Scout or something?) probably didn't help. I could kind of see using the Inquisitorial Cruiser and Deployment Shuttle on the player side, and the Prisoner Ferry would make a very neat hijacking target, but the super-expensive stealth fighter is probably less than useful.
So, overall impressions. Some of the book's chapters were poorly structured, and I'm probably not going to get all that much use out of the specific subsystems like the espionage mission game or the bounty hunting mission game. I do foresee adapting the general "You need X total effect in Y time" structure, so I am glad that this contributed to that development; I'm a little disappointed that they didn't realize they were doing to same thing several times and build a single unified structure for it, though, rather than reproducing it thrice with slight variants to fill page count. The Trust and tracking mechanics would both have been useful in the game I ran some springs ago, the careers were mostly good, and there were a fair number of things I could mine for a 40k Traveller game. I would argue that the content was not as broadly useful as that in Scoundrel, though - I wasn't hit with a "wish I'd had this" every chapter, and the rules subsystems, especially espionage, seemed sufficiently difficult as to be nearly unusable. Thus, a 3/5 for content seems appropriate - decent, but not noteworthy.
I was also pleasantly surprised to note a relative dearth of grammatical and editing errors - I believe there was one "your" in the careers section that should have been a "you're", and there was the missing column on the bounty payoff table, but other than those I was hard pressed to find clear errors. That I was critiquing structural issues in the espionage chapter suggests that Mongoose's editing went relatively well for Agent (either that or my copy from rpgnow already had errata and corrections applied). The art was mostly decent, with the career art being clearly superior to Scoundrel's, though the equipment art on page 103 seems rather... amorphous. I'm inclined towards a high 3 or low 4 out of 5 for production value and style here, though again lack of bookmarks is irritating in a pdf product.