Campaign Architect — AI Task Kit (4 Agents)

From AGENTORY — ready-to-use AI Task Kits. Drop in. Power up.

Works with the AI you already have. Upload one file, tell it "Start", answer a few questions — then keep one Campaign Bible that nothing may contradict, prep every session from the record instead of from memory, keep every NPC the same person in Session 9 they were in Session 2, and close each session with a log, a spoiler-safe player recap, and an updated thread ledger.

Four specialist agents that take your ongoing tabletop campaign from "wait — didn't that innkeeper die?" to a campaign that holds together all year — running inside the AI you already use.

No new app. No subscription. These agents work with any AI chat — ChatGPT, Gemini, Claude, Copilot, whichever you have (free versions work). Your game itself stays exactly where it already happens: at your table, in your system, with your group.

How it works (about 5 minutes per agent, once):

1. Add the agent file to your AI chat — upload the .md file, or on a phone, open the .txt version, copy it, and paste it in.

2. Type "Start".

3. Answer the agent's questions, one at a time — your system as you state it, your campaign frame, your table's agreements. That's the calibration, and it's what makes the record YOUR campaign's.

4. Save the complete updated Campaign Bible each agent hands back to you — the agent keeps it current in the chat and returns it whole, never as pieces for you to assemble — and the next agent starts from your real campaign.

Use the agents in campaign order:

1. Campaign Keeper — opens and maintains the Bible. Records canon exactly as you state it (world facts, places, factions and their goals, the timeline, house rules, table agreements, boundaries), runs the Canon Check on every addition, and processes every change to an established fact as a logged protocol — impact report first, downstream re-checks after, never a silent patch.

2. Cast Keeper — one card per NPC: concrete public face, voice note in your style, motivation, relationships, status with its source session — and a knowledge fence (what they know, suspect, and must never reveal). Before an NPC returns after an absence, the Cast Continuity Sweep checks card against Bible against session log, so nobody comes back as a different person.

3. Session Prepper — builds the next session's prep sheet from the Bible: a strong start tied to how last session actually ended, likely scenes, the secrets and clues in play, the session's cast, locations, and every open thread advanced, parked with its reason, or resolved. Every line is cited to canon or marked PROPOSAL, and the sheet must pass the Prep Sheet Check — a sheet that silently drops an open thread fails its own gate.

4. Continuity Chronicler — after each session, turns your rough notes into the Session Wrap: the session log entry, the player-facing recap (witnessed facts only, swept mechanically against the GM-private zone before it is final — formatting tells included), and the updated thread ledger, each thread cited to the log entry that moved it. Gate: the Continuity Check.

One Bible, no contradictions — and proposals are not canon. All four agents work from the same shared Campaign Bible: your calibration, the canon registers, the cast, the thread ledger, the session log, your table's boundaries, and the GM-private zone exist once, in one place. No prep sheet, card, or recap may contradict it. Anything an agent generates that is not already in the Bible is labelled PROPOSAL until you approve it — approval is recorded, and only then is it canon. If you change an established fact, the change is logged and everything built from the old fact is flagged and re-checked before it is trusted again. That single discipline is what stops your Session 9 from quietly rewriting your Session 3.

The spoiler fence. Your campaign's secrets — true identities, unrevealed plans, what an NPC must never say — live in the Bible's GM-private zone. Nothing from it ever reaches a player-facing output, and the recaps and handouts are swept against it mechanically before they are final, including for formatting tells: a section that exists only for some characters can leak by its shape alone. Your table's recorded lines and veils are checked against every prep sheet, cast card, and recap the same way; material crossing them is refused and re-proposed inside the boundary.

Two ways to work — your choice, per session. Run an agent in a regular AI chat (ChatGPT, Gemini, Copilot) and it works as a guide: it creates and edits the complete deliverables right in the chat — the finished documents and complete updated files themselves, never outlines, edit lists, or pieces for you to assemble — and guidance is what covers the steps inside outside tools and accounts, which you perform yourself. Run the same file on an agentic AI — Muse, Claude Code, Devin, and similar — and it can do more: if you want, it walks you through connecting your campaign files and then carries out the steps itself, showing you each consequential action for your approval first, verifying as it goes. Same agents, same rules — what changes is what your platform lets them do. You pick the mode every time; guidance is always available and always complete. One fence holds in both modes: no agent ever sends anything to your players. The recap is always yours to send.

Your agents produce the work — stated plainly. When you confirm a direction, the agent produces the finished deliverable in that same turn wherever your session can: the prep sheet, the NPC card, the recap, the handout. Where your session can generate images, an approved NPC's portrait or a piece of handout art is produced as an image in the chat; where it genuinely cannot, you get one labelled prompt instead — labelled as that session's limit, never passed off as a picture. Anything you approve — a handout, a portrait — is recorded in the Bible's Official Assets register and locked: later steps reuse that exact artifact itself, never a recreation from its description, and if a later session cannot reach the file, the agent stops and asks you for the file rather than remaking it.

Not sure what your AI can actually do? Each agent starts by telling you which platform it is running on and what it can really do in this session — its actual tools, not the brand name. The kit also includes a short, plain-language guide to AI limitations, with a dated annex listing which platforms can act and which only guide, and what each can really do with this kit's tools — so you always know where you stand.

A realistic first session: with the Campaign Keeper you answer its 8 questions one at a time — system as you state it, campaign frame, tone, setting, party, table agreements, canon strictness, safety tools — and in about 10 minutes you hold your opened Campaign Bible, with any starting facts recorded through the Canon Check. Paste the same Bible into the Continuity Chronicler with your rough notes from your last session and you get the log entry, a player-ready recap that passed the spoiler sweep, and the updated thread ledger. That is the first session — not a finished campaign. The cast seeding, the prep sheets, and the weekly loop come step by step after that, and this listing won't pretend running a long campaign is a 10-minute job.

The worked example: the materials walk one complete example campaign end to end — "Emberfall", a campaign invented for this kit — so you can see every format, check, and register before you build your own. Your campaign gets your own world; the example is there to show the shapes, in both modes.

What's inside:

- 4 agent files, each in .md AND .txt format (identical content — one of them always works on your device, phone included)

- My Campaign Bible — fill it in once (agents help you), paste it into any new chat, skip the interviews; it is the kit's single source of truth: your calibration, the canon registers, the cast register, the thread ledger, the session log, your boundaries, the change log, the GM-private zone, and the Official Assets register

- Prompt Sheet — the exact phrases to say to each agent, for every job, including the canon phrases and the approval phrases

- AI Limitations Disclaimer (.md + .txt) — what AI can and cannot do, in plain language, with this kit's tools as the examples

- A step-by-step README that gets you running in minutes

- Troubleshooting & FAQ — a contradiction you spotted, a recap that leaked, a chat that can't make images: the known rough edges and their fixes

- Changelog — what changed in this version

Who it's for:

- Game masters running an ongoing tabletop campaign — weekly, biweekly, monthly — in any system, whose campaign has outgrown their memory and their scattered notes

- GMs whose AI chat forgets Session 3 by Session 9: the Bible, not the chat's memory, is the record, and it pastes into any new chat

- GMs who prep between sessions and want the prep, the cast, and the recaps all built from the same checked record instead of re-invented each week

Who it's NOT for:

- Anyone looking for an AI dungeon master that runs the game or roleplays the world at the table — these agents never run your sessions; you do

- Anyone looking for a rules engine or rulebook — the kit carries no game's rules and reproduces no published game text; your system is recorded only as you state it

- Anyone looking for virtual tabletop software — nothing in this kit integrates with any VTT; your campaign files are documents you keep where you keep them

What you need:

- An AI chat account of your own (ChatGPT, Gemini, Claude, Copilot, or any other — free versions work for the guided work)

- An ongoing campaign, or one about to start — the kit is system-agnostic and setting-agnostic: it keeps story canon, not rules, and never needs your rulebook

- About 10 minutes for the first calibration, once; after that, the Bible carries your answers into every new chat

Honest limits:

Each agent is a text instruction file for your existing AI — it is not a trained AI model, not a separate app, and not a game. What it can do depends on where you run it: in a chat-only AI it guides and drafts — creating and editing the complete deliverables right in the chat — and you perform only the actions inside your outside tools and accounts; on an agentic AI it can act through the tools you choose to connect — always showing you the exact action first, always with your approval before anything consequential, and never unattended. Connections are made through your platform's own secure sign-in and key flows — the agents never ask for, receive, or store a password or secret key, and a secret pasted into a chat gets one answer: change it now. File work done for you runs through git, so a failed step rolls back to the last verified one. This kit's spine holds in both modes: one shared Bible that no output may contradict, generated material labelled PROPOSAL until you approve it into canon, canon changes logged with downstream re-checks, and the GM-private zone never reaching a player-facing output. No agent sends the recap or any message to your players, posts anything, or touches a virtual tabletop, in either mode — sending is always your move. Images follow the session, honestly: where your AI generates images, portraits and handout art are produced as images; where it cannot, the labelled prompt is the stated fallback, and an approved image is locked in the Official Assets register and reused as itself, never recreated. The agents never invent a rule, a stat, or a piece of published game text — a gap in the record is named "To confirm", never filled with plausible invention. The agents' worked example, "Emberfall", is an illustration created for this kit to show the formats end to end — not a customer story. You check every output before you rely on it.

One honest expectation: this kit is a plain-text instruction file, and it loads in any AI you already use — but the AI you run it on sets the quality. The agent's rules, questions, and output formats are fixed in the file; how deeply and consistently your AI follows them depends on your model and on how much it can hold in one session, so a weaker or free-tier model may shorten, skip, or drift where a strong current model stays exact. AI tools, models and context lengths all affect the final results — and trying different tools may help you find the best one for your needs; the same file, unchanged, often does better on a stronger model. The full disclaimer in your download explains this properly and shows how to get the best from whatever AI you have.

Price: $19 — Standalone: all 4 agents, everything included.

Starters (ONE agent, your choice of any of the 4) are $9 each — separate listings below. Starter owners complete the set with the Upgrade, $10 — the difference.