I enjoy widebody aircraft. When high-fidelity long-haul aircraft like the PMDG Boeing 777 land in Microsoft Flight Simulator or Prepar3D, or the FlightFactor 777 for X-Plane, most simmers face a common dilemma: we want to fly complex widebody jets based on real-world flights, but we don't always have 10 to 14 hours of free time.

While the 777 is famous for ultra-long-haul intercontinental routes, real-world airlines routinely operate it on 2- to 3-hour legs—whether for high-density regional routes, cargo runs, or passenger-bookable repositioning flights. I find that I either have to sift through paywalled flight-tracking databases or spend an inordinate amount of time and effort finding these shorter real-world flights in specific geographic regions.

To solve this, I created a lightweight, token-based "Prompt Agent." You can copy and paste this prompt directly into ChatGPT, Gemini, Claude, or any AI to instantly generate a tailored table of real-world routes for your next flight simulation session.

If you ask an AI a generic question like "Give me some short 777 flights in North America," you often get generic, imprecise, or hallucinated flight numbers.

To get reliable, deterministic flight plans that match real-world operations, the AI needs strict boundaries:

1. Specific Aircraft Variants: Distinguishing between 777-200ER, 777-200LR, 777-300ER, and Freighters.

2. Geographic Constraints: Filtering by departure or arrival regions (e.g., North America).

3. Sorting & Filtering: Ordering by flight distance or departure airport so you can quickly pick a flight that fits your schedule.

4. Fallback Calculation: If schedule times aren't publicly indexed, calculating realistic block times based on standard 777 cruise speeds ($\text{Mach } \sim 0.83$) and taxi estimates.

At its core, the prompt tells the AI model: "Look at the command string A$ M$ G$ and the region header, look up what those variable strings mean in the definitions below, and generate a standardized table based on those exact rules."

Here is the step-by-step literal translation of what the AI reads and executes:

1. Parse the Command Line:

   • Select aircraft set A$ (all 777-200, 777-300, and 777 Freighter variants).

   • Apply Departure From Filter: North America to restrict origin airports.

   • Set primary sort to M$ (Distance in nautical miles, ascending).

   • Set secondary sort to G$ (Departure ICAO code, ascending).

2. Retrieve Flight Data: Find up to 50 active or recently operated real-world flights matching these criteria, ensuring a mix of short, medium, and long-haul sectors.

3. Calculate Fallbacks: If a flight's official block time isn't publicly indexed, estimate it using standard 777 operational parameters ($\text{Mach } \sim 0.83$ cruise speed + 15–25 minutes ground taxi) and label the source as "Inferred". Otherwise, mark as "Documented Example".

4. Format the Output: Render the final results into a structured 12-column Markdown table using exact token field names ($D\$ - O\$$) without extra commentary.

Rather than writing a long, verbose paragraph every time you want a route recommendation, this prompt uses a tokenized command header. This command header uses replaceable strings. Think of $A\$$ as "A string" of text, $B\$$ as "B string", etc.

A$ M$ G$
Departure From Filter: North America

Breakdown of the Logic:

• Command Header (A$ M$ G$): Functions like passing parameters into a function.

  • A$ selects the aircraft set (All 777 Passenger & Freighter variants).

  • M$ sets the primary sort key to Distance (nmi) in ascending order (shortest flights first).

  • G$ sets the secondary sort key to Departure Airport ICAO.

• Filter Parameter: Departure From Filter: North America restricts all results to flights departing from North American airfields.

• Token Definitions: Quick references ($A\$ - O\$$) that declare exact data types and table columns.

When you execute the prompt, the AI skips any conversational fluff and generates a clean, structured table sorted directly by distance:

When I was designing this prompt, my key goal was keeping it as simple and user-friendly as possible.

You could build a hyper-complex prompt that accepts any aircraft type, any route distance threshold, and multi-region filtering parameters. However, increased prompt complexity often leads to higher user friction, unexpected AI interpretations, or missed instructions. Nevertheless, I encourage you to experiment with the prompt.

Known Limitations of This Design:

• Aircraft Tokens Are Hardcoded: The token set $A\$, B\$, C\$$ is specifically calibrated for the Boeing 777 family.

• Manual Parameter Tweaks: Changing to a different airframe (e.g., Airbus A320 or Boeing 737) requires editing the token definitions.

By accepting these minor limitations, the prompt remains lightweight, fast, and 100% reliable for quick copy-pasting into any LLM context.

For the casual flight simmer, this article may look technical and even complex, but for the serious simmer and those who're willing to follow simple instructions, using the prompt is quite easy. In a nutshell, I explained briefly how the prompt operates: you just copy the prompt and paste it into your favorite AI.