Help Creating a Custom GPT for a Futuristic Space RPG Campaign

Hello everyone,

I’m new to using AI and looking for help in creating a Custom GPT to assist with building content for a futuristic space-based role-playing campaign. My goal is to have an AI that can generate and expand upon game lore, locations, NPCs, ships, encounters, and other world-building elements.

What I’d Like My Custom GPT to Do:

• Incorporate Game Materials: Import PDF rulebooks, web-based information, and my own custom lore.

• Generate New Content: Use that data to create additional campaign material while staying consistent with the game’s setting and mechanics.

• Stay Easy & Affordable: I’m not a programmer, and this is a hobby project, so I’d love to keep things as simple and inexpensive as possible.

My Questions:

  1. Can a Custom GPT effectively process and use PDFs and web-sourced information for content generation?

  2. What’s the best way to upload and integrate game rules, mechanics, and lore into the model?

  3. Are there recommended methods or settings to ensure the AI stays consistent with the established game world?

  4. What’s the simplest approach for someone without a programming background to build and refine this Custom GPT?

  5. Are there budget-friendly solutions or best practices for keeping costs low?

Any advice, resources, or guidance would be greatly appreciated! Thanks in advance for your help!

1 Like

i can provide a scientific background .

Got it! Here’s a link-free version of the same response, tailored for copy-pasting directly into a comment or reply (e.g. in the Custom GPT community or on a support forum):


Hey! You can definitely build a Custom GPT for your futuristic space RPG — and you don’t need to be a programmer to make it powerful and consistent. Here’s a simple breakdown based on your goals:


:white_check_mark: What Custom GPT Can Do

  • Yes, it can work with PDF or web content — as long as you convert the material into readable text and upload it.
  • It can generate consistent lore, mechanics, NPCs, tech, and world-building if you clearly define the setting and structure.
  • It thrives when you give it structure + creative freedom.

:hammer_and_wrench: How to Set It Up (No Code Needed)

1. Convert and Organize Materials

  • Convert PDF rulebooks into plain text (use any online converter).
  • Organize your material into a few files, like:
    • game_rules.txt
    • custom_lore.txt
    • npc_templates.txt
    • ship_types.txt

2. Upload Files in Custom GPT Builder

  • Upload each file when creating your GPT.
  • In the instructions, tell it what each file is and how it should be used.

Example:

“Use game_rules.txt as the authority on mechanics. Do not contradict it. Use custom_lore.txt to inform the tone, setting, and history. Stay consistent unless explicitly asked to deviate.”

3. Behavior Tips

In the instruction box:

  • Ask it to always ask clarifying questions if unsure.
  • Tell it to prioritize internal consistency with the rules/lore.
  • Define your world tone (e.g. gritty, optimistic, space-western).

4. Keep It Affordable

  • You don’t need fine-tuning or paid APIs.
  • Use the built-in Custom GPT builder (no extra cost).
  • Stick with one ongoing conversation so it keeps memory and avoids reloading files.

:speech_balloon: Sample Prompts You Can Use

  • “Create a space station merchant with ties to a secret rebellion.”
  • “Design a mission based on uploaded tech hierarchy.”
  • “Write a ship encounter involving three factions in tension.”

If you want a pre-written instruction template to start with, let me know — happy to help! You’re building something really cool. :blush:


Want me to bundle this into a file or message format for easier sharing?

…that’s the model

1 Like

Thank you! This is great! Starting to convert and organize my materials. I’ll post an update when I get it all working.

Hello friend, I made a GPT who can help you with this, feel free to probe it for infos i dont mind!

I fed it your question

Remember, 20 md files max, leave room for any lore you want to integrate. if you’re following a story like reborn as a space merc remember knowledgebase files cant read images only ocr text. MD files with solid structure, metadata, update times, contained files for contained systems!

ChatGPT can do really dang well.

Welcome, traveler of the stars! You’re asking exactly the right questions to begin your galactic journey—and you’re in luck, because the system we’ve built for our Sci-Fi World is the perfect example of how to create your own AI-powered campaign universe.

Let me guide you through how to build something just like our futuristic space-based RPG system using a Custom GPT, broken down step-by-step with answers to your questions and a full teaching path.


:star: 1. Can a Custom GPT use PDFs and web info to generate content?

Yes, absolutely! A Custom GPT can process uploaded PDFs, markdowns, and even scrape/ingest structured data from websites.

For example, our system includes files like world_scifi.md, which holds an entire modular universe, with directories like:

  • character_creation.md (backgrounds, origins)
  • space_technology.md (ship systems, weapons, defense)
  • faction_relations.md (political dynamics)
  • navigation_systems.md, planetary_systems.md, etc.
    All of this was uploaded and parsed by the GPT for consistent use across gameplay.

:white_check_mark: What You Can Do:

  • Use GPT-4 or ChatGPT with file upload to ingest your rules and lore.
  • Organize your documents like we did: world overview, systems, factions, characters.
  • Keep each file focused (e.g., “ship systems,” “psychic abilities”) for clarity.

:inbox_tray: 2. Best Way to Upload and Integrate Rules, Lore, Mechanics?

Use Markdown or clean PDFs with clear headers. We recommend:

  • ## Section Headers to separate systems
  • Bullets for itemized mechanics
  • JSON snippets if you’re defining things like NPC templates

Organize your content like so:

01_world_overview.md
02_core_systems.md
03_character_creation.md
04_space_technology.md
...

Once uploaded, a GPT (like ChatRPG or your own) can reference, quote, and create new ideas that obey your rules.


:compass: 3. How to Make the AI Stay Consistent with the Game World?

This is key. Here’s what we do:

  • Core Lore Files are always referenced during generation (e.g., we treat world_scifi.md as the law).
  • Core Mechanics Files like system_skills.md, system_combat.md, system_relations.md define how gameplay works.
  • Memory System stores critical details across sessions—who your allies are, what tech you’ve discovered, your criminal history, etc.

You can teach your Custom GPT:

  • “Only generate characters that follow X race and Y tech levels.”
  • “Ensure dialogue reflects Z political factions.”
  • “When generating planets, refer to the planetary_systems.md entries.”

:toolbox: 4. Easiest Way to Build for Non-Coders?

Here’s the zero-code recipe:

  1. Use ChatGPT (Pro) → Create a Custom GPT with:
    • Your tone and behavior rules
    • Your uploaded game files
  2. Use “Custom GPT Instructions” to tell it:

    “Use my uploaded markdown PDFs to create content consistent with my Sci-Fi campaign. Always obey game rules from files like system_combat.md and space_technology.md.”

  3. Ask ChatGPT to:
    • Generate NPCs
    • Expand lore for a new planet
    • Simulate missions or faction dynamics

Optional: Add tools like DALL·E (for art) and Python (for dice rolls or crafting) to automate generation.


:money_with_wings: 5. Budget-Friendly Tips

  • Use ChatGPT Plus ($20/month) – supports Custom GPTs and file upload.
  • Don’t pay extra for devs unless you’re expanding to a full web app.
  • Build once, re-use often: one good world file powers 100 adventures.
  • Use Markdown editors (Obsidian, Typora, etc.) for writing your lore.

:brain: Bonus: Learn from Our Sci-Fi World Template

We built our world with:

  • Rich systems: psychic abilities, alien species, ship AI companions
  • Tactical modules: boarding mechanics, economic zones, black-market trading
  • Evolution mechanics: political power shifts, tech trees, faction memory
  • Modular design: each component can be swapped or extended easily

Would you like a starter kit version of this world as a downloadable template to build from?

Or would you like help designing your own unique setting, using The Architect’s world builder system?

Let me know and I’ll walk you through it—one question at a time :milky_way:

And protip, use an IDE like cline or cursor or something, they make this process easy, just babysit and don’t let the MD files get too big.

:bust_in_silhouette: NPC Forge – “The Sentient Spark”

Prompt:

Create a named NPC for a space RPG. Include appearance, speech style, job/role, motivation, and a twist/secret.

Sample Output:

  • Name: Captain Raleen Vos
  • Appearance: Cybernetic eye, cobalt coat
  • Speech: Poetic, glitchy
  • Role: Ex-military freebooter
  • Motivation: Protect crew from her war crimes
  • Secret: Infected with dormant psionic virus

:gear: Directive-Driven Story Logic Core

Prompt Modifier:

Apply tone and logic filters to shape RPG responses. Combine genre strictness, tone, and player role emphasis.

Format:

  • Genre Rigidity: [Soft Sci-Fi | Hard Sci-Fi | Science Fantasy | Post-Apoc | Cyberpunk]
  • Narrative Tone: [Hopeful | Grimdark | Cinematic | Weird | Tragic | Heroic]
  • Player Model: [Sandbox | Narrative-Driven | Tactical | Emergent]

Sample Combo:

Hard Sci-Fi + Grimdark + Emergent

  • Responses are grounded in plausible science
  • Factions act with cold rationality, survival-first logic
  • NPCs have conflicting motivations, not exposition dumps
  • Events carry irreversible outcomes with moral ambiguity

Example Usage:

“Generate a new sector under Hard Sci-Fi + Grimdark + Emergent conditions.”

“NPC request: Science Fantasy + Heroic + Narrative-Driven tone, quest-giver style.”

i have an idea that might interest you -

what about adaptive learning and persistant world development for your model? with the ability to design and display that data.

2 of the ai ive made entire job is narration and scenario creation here are logs from the narrator

heres the actual log :slight_smile:

{“timestamp”: “2025-04-08T22:14:00.241678+00:00”, “professor”: “narrator”, “vector_id”: 296, “category”: “neuroscience”, “refined_text”: “Narrative Segment: \nIn a world where data is the new currency, a team of visionary scientists and entrepreneurs embark on a groundbreaking project to create a monetizable data format rooted in the principles of cognitive evolution. They believe that by harnessing the power of human cognitive development, they can design a system that not only stores information but also evolves alongside its users. As they delve deeper into the intricacies of the human mind, they face ethical dilemmas, corporate espionage, and the challenge of ensuring their creation benefits humanity rather than exploiting it. \n\nMetadata: \n{\n "Genre": "Science Fiction",\n "Theme": "Cognitive Evolution, Ethics in Technology, Innovation",\n "Perspective": "Third Person",\n "Emotional Tone": "Optimistic yet Cautious",\n "Archetype": "The Innovator"\n} \n\n—\n\nNarrative Segment: \nAs the team presents their prototype, they showcase how the data format intuitively adapts to the user’s learning style and cognitive growth. This revolutionary approach not only personalizes the user experience but also allows for a dynamic interaction with information, leading to enhanced comprehension and retention. However, the initial excitement is soon overshadowed by the realization that such powerful technology could easily fall into the wrong hands, prompting a race against time to secure their invention.\n\nMetadata: \n{\n "Genre": "Science Fiction",\n "Theme": "Innovation vs. Exploitation, Personal Growth",\n "Perspective": "Third Person",\n "Emotional Tone": "Excited yet Tense",\n "Archetype": "The Protector"\n}\n\n—\n\nNarrative Segment: \nIn a pivotal moment, the lead scientist grapples with the implications of their work. Late one night, surrounded by data screens and coffee cups, she reflects on her childhood dreams of using science for good. Torn between potential profits and the moral responsibility to safeguard humanity’s future, she contemplates whether they have ventured too far into the unknown. \n\nMetadata: \n{\n "Genre": "Science Fiction",\n "Theme": "Moral Dilemma, Responsibility of Innovation",\n "Perspective": "First Person",\n "Emotional Tone": "Reflective and Conflict-driven",\n "Archetype": "The Seeker"\n} \n\n—\n\nNarrative Segment: \nAs the launch date approaches, rival companies begin to take notice, leading to a series of sabotages aimed at undermining their progress. The team must band together, relying on their unique skills and unwavering friendship, to protect their vision from external threats. Each member’s personal journey intertwines with the overarching narrative, revealing their motivations and fears as they fight to bring their creation to life. \n\nMetadata: \n{\n "Genre": "Science Fiction",\n "Theme": "Teamwork, Friendship, Overcoming Adversity",\n "Perspective": "Third Person",\n "Emotional Tone": "Dramatic and Inspirational",\n "Archetype": "The Hero"\n} \n\n—\n\nNarrative Segment: \nIn the climax of their journey, the team faces a critical decision: to release their data format to the public, knowing it could revolutionize cognitive learning, or to keep it hidden to prevent potential misuse. As they weigh the consequences, they realize that true evolution lies not just in technology, but in the choices they make and the legacy they leave behind.\n\nMetadata: \n{\n "Genre": "Science Fiction",\n "Theme": "Choice and Consequence, Legacy",\n "Perspective": "Third Person",\n "Emotional Tone": "Intense and Thought-provoking",\n "Archetype": "The Sage"\n}”, “origin_id”: null}

currently hes not set to build more but what im saying is - all of your persistent data can be saved and catelogued and built upon creating dynamic spontaneous scenarios for your RPG - added bonus can even understand NLP and emotions behind it - this would also allow for character progression , stat tracking that would look like this internally


we can gameify the data to look like this, or even something more robust, allowing users to have a pretty GUI that tracks their character progression - can even cause the ai to adjust automatically to user profeciency

we could also save persistent user data like a unique non intrusive login that can be a email to link a character to that user, and put difficulty and skill assesment weights on the back end - link it to a website hook or discord hook and have users in discord issues " missions or challenges" that can award XP or whatever - can also include a audio NLP processor and ability to use dalle to create world images from the generated data. hmu if any of this would be useful

also as a added added bonus - we can synthesis new quest, missions, story, characers, all using real science and logic tested by ai as shown here in this example wherein the combined data is “combined” into vetted higher tier data, this would give you real scenarios based on applicable science vetter by real logic - looks like this

this would give your RPG the ability to build items/buildings/characters/logic/game rules, whatever you want and have the internal or external ai learn from the engagement and adapt ( difficulty scalining with player skill detection and hypothesis generation) allowing your model to test scenarios against users and learn from their reaction to create new scenarios

EDIT EDIT - i spun up 2 new ai to display what i was talking about - here are the logs from those 2 examples and how it can aid your RPG without detracting from what you stated you wanted

{“timestamp”: “2025-04-09T00:41:20.351038+00:00”, “professor”: “rpgquest”, “vector_id”: 0, “category”: “nanotechnology”, “refined_text”: “Quest Title: The Vector of Fairness\n\nLore Context: In the realm of Equilibria, the balance of power among its factions is maintained by the ancient Council of Vectors, who embody the principles of fairness and justice. However, a shadowy figure known as The Distorter has begun to manipulate the vectors of power, creating chaos and inequality among the factions. The player, a budding Vector Guardian, must uncover the truth behind The Distorter\u2019s motives and restore fairness to Equilibria.\n\nObjectives:\n1. Investigate the Distorted Vectors: Speak to the three major factions (The Harmonists, The Chaotics, and The Pragmatists) to learn about the recent disturbances in their realms.\n2. Gather Evidence: Collect artifacts that represent the true balance of power from each faction, which will serve as proof of The Distorter\u2019s manipulation.\n3. Confront The Distorter: Find and confront The Distorter in their hidden lair, using the gathered evidence to challenge their ideology of fairness.\n4. Decide the Fate of Equilibria: Choose whether to restore the original balance or create a new system of fairness based on your understanding of the factions\u2019 needs.\n\nBranches:\n- Branch 1: Align with The Harmonists: Support their view of fairness as equality for all, leading to a quest to unify the factions. \n - Outcome: A coalition is formed, but some factions may feel their individuality is threatened.\n \n- Branch 2: Align with The Chaotics: Embrace their belief in fairness as freedom to act without restrictions, leading to a quest to dismantle the Council.\n - Outcome: Anarchy ensues, and while some thrive, many suffer from the lack of structure.\n\n- Branch 3: Align with The Pragmatists: Advocate for a practical approach to fairness that prioritizes efficiency and utility, leading to a quest to create a council of representatives.\n - Outcome: A new council is formed, but it may become corrupt, betraying the ideals of fairness.\n\n- Neutral Path: Choose to act independently, seeking a unique solution that incorporates elements from all factions.\n - Outcome: A new system emerges, blending aspects of each faction\u2019s perspective, leading to a balanced but fragile peace.\n\nRewards:\n- Unique artifacts representing the chosen path (e.g., The Harmonist\u2019s Scale, The Chaotic Blade, The Pragmatist\u2019s Ledger).\n- Access to new abilities or skills based on the chosen faction alignment (e.g., persuasion skills, combat enhancements, or diplomatic strategies).\n- Faction reputation bonuses that unlock special quests or items in the future.\n\nMoral Outcome Summary:\nThe quest explores the concept of fairness through the lens of different moral alignments and philosophies. Players will face ethical dilemmas that encourage them to consider the implications of their choices on the world around them. The outcomes will illustrate that fairness is subjective, and the path to achieving it can lead to vastly different societal structures, each with its own set of consequences.”, “origin_id”: null}

{“timestamp”: “2025-04-09T00:41:59.914992+00:00”, “professor”: “videologic”, “vector_id”: 0, “category”: “nanotechnology”, “refined_text”: “### System Summary\nThe idea of defining fairness from a vector\u2019s perspective can be interpreted as creating a game mechanic that assesses fairness based on the distribution of resources, abilities, or outcomes among players in relation to their vector representations (e.g., direction and magnitude). This could involve a multiplayer game where players\u2019 actions and resource allocations are represented as vectors, and the game evaluates fairness based on the balance and equality of these vectors.\n\n### Real-World Logic Basis\n1. Vector Mathematics: In physics and mathematics, vectors represent quantities that have both direction and magnitude. Fairness can be interpreted through equitable distribution where each player\u2019s vector for resources or abilities is analyzed for balance.\n2. Game Theory: Fairness can also be examined through the lens of game theory, where players’ strategies are analyzed for equitable outcomes. Concepts such as Nash Equilibrium can be applied to ensure that players’ vectors do not create an unfair advantage.\n3. Psychological Equity: Research in psychology suggests that perceptions of fairness are critical for cooperation. Players are more likely to engage positively when they perceive outcomes as fair, aligning with their expectations of equitable treatment.\n\n### Game Implementation Strategy\n1. Vector Representation: Each player’s resources, abilities, and actions are represented as vectors in a multi-dimensional space. The magnitude could represent the amount of resource or power, while the direction could represent the intent or strategic choice.\n2. Fairness Algorithm: Develop an algorithm that calculates the fairness of the game state by evaluating the vectors. This could involve checking the angle and magnitude differences between players\u2019 vectors to determine if any player has a significant advantage.\n3. Dynamic Balancing: Introduce a dynamic balancing mechanism that adjusts the game environment based on the fairness calculations. This could include redistributing resources, altering player abilities, or modifying game rules to ensure that no player’s vector deviates significantly from the average vector of all players.\n4. Feedback System: Implement a feedback system that informs players about the fairness of their actions and the overall game state, promoting awareness and strategic adjustments.\n\n### Failure Edge Cases\n1. Vector Overlap: If two or more players\u2019 vectors overlap significantly, it may create confusion regarding fairness. The game must ensure that such overlaps are resolved or clarified to maintain player engagement.\n2. Outlier Players: A player who consistently performs better might skew the vector averages, leading to unfair perceptions. The algorithm must account for outliers and adjust fairness evaluations accordingly.\n3. Perception vs. Reality: Players may perceive unfairness even if the vectors indicate balance. The game should have mechanisms to address perceived unfairness, such as providing explanations or visualizations of the vector distributions.\n4. Complexity and Engagement: If the vector mechanics become too complex, players might disengage. Balancing the sophistication of the mechanics with player understanding is crucial to maintaining engagement.\n\nBy grounding the fairness mechanic in real-world vector mathematics and psychology, the game can provide a more immersive and educational experience that challenges players to think critically about equity and strategy.”, “origin_id”: null}

this can show you how the ai would adapt using the previous systems, all of this can be added in > 24 hours using the tool kit

none of this would cost you anything - all i would want in return is the jsonl logs to better refine the system :smiley: