There is a question that closely connects imagination of both players and game designers: will a game feature X, Y, and Z? Naturally, games cannot represent all complexities of the world (even Dwarf Fortress does not have seemingly straightforward items like boats) - many of them have to be abstracted away to create an approximation, a model.
As mathematicians say, "all models are wrong but some are useful". In this context, useful should probably mean logical, interesting, challenging, sometimes fun (sometimes - fun is not the only emotion evoked by good games). This is reflected by the hotly debated issue of mana points in grand strategy games. It is also an important culprit in the hunt for sins of espionage systems in strategy games, since many of them (with honourable exception of hooks in Crusader Kings 3) model espionage as knowledge tax where you just pay to uncover something. How interesting is paying taxes?
In attempt to build a different espionage system, one founded on constructive actions and discrete results, Espiocracy models espionage as a world full of secrets and opportunities, which have their inherent dynamics, can be acquired and weaponized, and may even bite you back.
Secrets are one of the many facets behind reactive world building in the game. Murders happen, people can get away with them, but the act will stay in memory of the perpetrator as a secret. These memories can be elicited from friends, in drunken conversation, or during interrogation. Sometimes they expand to the larger world of witness accounts and hard evidence (photos, recordings, signed documents), giving modus operandi of removing witnesses and Hoover-style kompromat hoarding.
Their inherent design is simulationist. Real world secrets are surprisingly granular and escape any wider generalization. There is an obvious angle of significant law breach - but what happens when the law is unjust or is not enforced or is dropped but the stigma remains? The UK prosecuted "homosexual acts" in the 50s and convicted GCHQ's Alan Turing to chemical castration but Guy Burgess, Soviet spy in MI5, who "made no attempt to conceal his homosexuality" did not have any issues. The law was dropped in 1967, but the British government in its system of "positive vetting" rejected homosexual candidates all the way to 1991. It is anything but simple!
The game tries to capture most important parts of this granularity by making secrets local, personal, and situational. In the USSR believing in communism isn't a secret, in the USA it is. For one person, love affair is dramatic secret, for another it's a reason to boast.* Murders during the war are much less of a secret in contrast to a long period of peace. Above all, this leads to continuous instead of binary (secret or non-secret) spectrum, which nicely plays into metaknowledge about secrets: falsifications, allegations, and different types of evidence weigh differently on the impact.
Speaking of which, secrets serve as a full blackmail material - not only in the form of threat but also as a real tool with large impact. Leaking a secret to the press can destroy career, purge an organization, or even topple governments. It's not limited to public scrutiny and can be used to open doors, for instance revealing a secret to the government can give you green light and special funding to harass the organization out of your country.
Player, as an actor in the game, is also subject to the world of secrets. Since you are an intelligence agency, rules are made to be broken - by accepting the challenge stemming from a new secret in your backyard, the risk of blackmail, whistleblowers, and scandals.
Coming back to the introduction and murder example, people are often jokingly asking if you can kill your own president in the game - that's the place where system of secrets kicks in. Following intelligence rule of need-to-know, this secret is privy only to operatives engaged in the operation. It arises on the first day of planning and evolves over time ("planned to kill president" has obviously different weight than "killed the president"). Successful assassination, even with the best cover in the world, can still leave traces of evidence for nosy reporters, detectives, and other players. Moreover, it remains in the memory of executing operatives. Should they stay in the conspiracy circle or be eliminated? If elimination goes partially wrong, won't some of them run for their lives and reveal the secret to the world? Maybe blaming everything on one scapegoat is enough? (Here's also where the stakes are spiced up by one of the game over conditions: grand treason. If this particular secret is revealed, it can lead to complete dissolution of intelligence apparatus, loss of all operatives, and therefore end of the game.)
There are many less grave secrets: illegal wiretaps, enhanced interrogation, creative accounting, deals with gangs and terrorists, experiments on humans and animals (that comes down to mentioned local sensitivies; in the era of CIA's illegal experiments on humans in MKUltra, the idea of eavesdropping "Acoustic Kitty" was feared as too dirty for people concerned about animals), and so on. There are even rare secrets that can be brought upon the player by operatives, such as covering up a stupid crime commited by an operative too useful to be imprisoned.
Following the KISS principle (keep it simple, stupid!), Espiocracy features literal extensions of player agency: opportunities. Examples include ability to intecept a person in transit, to infiltrate an organization with an agent of ideal background, to steal a piece of technology presented at an expo, to funnel money from government to an actor, and so on. They are essentially a discrete currency of the intelligence world, acquired during operations, bought from actors and other players, received from the government, or sometimes just randomly stumbled upon.
On the most basic level, they are as intuitive as the word "opportunity" can get (which is why this section is rather short). From the strategic point of view, they introduce one more layer of decisions that contributes to planning, preparation, cooperation, coordination, and a few other staples of decision making. Some of the opportunites get more complex dynamics, limiting their use to a particular time window, location, or a set of requirements (the last one can be broken by getting, for instance, "creative accounting" secret). Other times, they function as a sincere favor system - after assisting local mafia, they may return the favor in the form of opportunity. There are also cases where opportunities can make failure worthwhile by opening new opportunities. Rarely, some opportunities can feed into the paranoic side of the game, where a player can dangle manufactured opportunities to ambush other players.
"In chaos, Madame Ambassador, there is opportunity" - CIA officer Douglas London