Start by mapping the city’s infrastructure into clear layers—traffic, utilities, communication, security. Each layer should have a small, well‑defined interface: a login prompt, a checksum check, a routing table. For realism, use real‑world protocols in a stripped‑down form; for example, a simple key‑value store for traffic lights, a mock TLS handshake for communication, and a hash function for passwords. Make the puzzles incremental. The first hack could be a brute‑force login that reveals a weak password; the next step is to exploit a known buffer overflow, but you’ll have to locate the exact memory address in a hex dump that’s been rendered in the UI. Keep the code snippets minimal—just enough lines to be solvable without scrolling through hundreds of lines. Introduce time pressure: the system will trigger a counter‑measure after a set number of attempts, so the player must think fast. Add a narrative consequence that is visibly tied to the code they write—e.g., the traffic lights go green, the mainframe logs a new entry, a character’s dialogue changes. That immediate feedback keeps the puzzle grounded in the story. Balance difficulty by allowing a “debug” mode that lets you see variables and memory, but make it optional or require a separate skill to access. That way, a player who wants to experiment can, but the core experience remains a test of logic and pattern recognition. Finally, give each puzzle a unique flavor: a street‑level hack that uses a small script, a corporate breach that requires reverse‑engineering a binary, and a city‑wide takeover that demands orchestrating several sub‑hacks in parallel. Variety keeps the gameplay fresh while still feeling like a cohesive hacking adventure.