Here's the system: [hider=My Hider] [b]1. The Ledger — Core Oracle Mechanic[/b] You need a Ledger to somehow record tokens of four colors. Standard Ledger composition: 4 Aligned, 4 Opposed, 2 Twist, 1 Root (11 tokens total). Drawn tokens go into a discarded Ledger, spending it, only refreshing when a scene ends (see 1.4). This means the Ledger depletes over a scene, and the flavor of your fiction shifts as it does: a Ledger running low on Opposed tokens feels like momentum is with you. [b]1.1 Asking the World a Question (Oracle Use)[/b] When fiction is ambiguous and you need the world to answer ("Is the safehouse being watched?" "Does the Association already know my name?"), state the question as a yes/no, judge its Likelihood, adjust the Ledger, draw one token. Questions regarding the world have their own separate Ledger pool to draw from. [hider=My Hider] Likelihood - Adjustment before the draw Certain - Remove all Opposed Likely - Add 2 Aligned 50/50 - No change Unlikely - Add 2 Opposed Impossible - Remove all Aligned [/hider] [b]1.2 Taking Action (Task Resolution)[/b] When a character attempts something with real stakes — casting a spell, sneaking past an someone, reading a bounded field, cracking a family cipher — draw a pull of tokens equal to the relevant Rating (see Character Creation, capped 1–5) from the Ledger, all at once, and resolve by majority color: This pull's tokens go to the discard pile, depleting the Ledger for the rest of the scene. [b]1.3 Contested Actions[/b] When two intentions directly conflict (your Reinforcement-strengthened block vs. another writer's Dead Apostle claws), both writers draw simultaneously and post both results in the open before either narrates the outcome — this prevents one party from writing the other's loss (or their own win) before the dice, so to speak, have actually landed. Whoever has more Aligned tokens (Root counts double) wins; ties favor the defender. Whoever wins narrates the outcome for their own character only; the other writer narrates their own character's reaction. Nobody writes another writer's character taking damage, being hit, or losing an item without that writer's own follow-up post confirming it. [b]1.4 Refilling the Ledger[/b] The Ledger refills (all discards returned) at the start of each new scene — in this context, a one clearly-bounded sequence of posts (a heist, a conversation, a fight) — or immediately whenever it runs completely dry mid-scene. [b]2. PRANA & OD — The Resource Layer[/b] Separate from Ledger tokens, each mage has a personal pool of Prana Tokens representing stored Od plus environmental Prana drawn through their Circuits. Prana Pool size = Circuit Quantity rating (see 3.4), refreshed to full after a full scene of rest, meditation, or access to a leyline/workshop. Every Skill use or Minor Mystery (Section 4) has a Prana Cost paid before you draw your action pull. Paying it is not optional and is not refunded on failure. Running out of Prana doesn't stop you from acting, it means you’re drawing from a nearly-empty tank. Any spell cast at 0 Prana costs 1 point of Strain instead (track separately; 5 accumulated Strain drops you unconscious/collapsed, mirroring burnout. [b]3. CHARACTER CREATION — Building a Mage[/b] Work through these steps in order. Every Rating is 1–5 and equals the number of tokens you draw for related pulls. [b]3.1 Concept & Archetype[/b] Pick or write one: Clock Tower researcher, disowned family heir, self-taught mage, Church-affiliated individual, Dead Apostle (non-Ancestor) hunter, itinerant puppeteer, alchemist, curse-broker, Atlas Institute washout, mage outside the Association's reach (custom.) [b]3.2 Origin & Alignment[/b] Drive: a core truth about the character's soul. Your Drive grants a permanent +1 token to any pull thematically aligned with it, and a GM-less "temptation" clause: acting against your Drive's nature always draws from a Ledger with one Opposed token added, unremoved. [b]3.3 Magic Circuits[/b] Quality (1–5): rank from Poor to Excellent. Sets your ceiling — you cannot use a Skill at a Difficulty exceeding your Quality without accepting 1 automatic Strain. Quantity (1–5): sets your Prana Pool size and how many active/ongoing spells you can sustain at once. [b]3.4 Foundation (School of Thaumaturgy)[/b] Choose two schools of magecraft to specialize in, called Foundations. Anything from Alchemy, Siberian Shamanism, Kabbalah to all sorts of other things from myth and academia can be Foundations. [b]3.5 Mundane Skills[/b] Rate 1–5 each: Investigation, Combat, Social, Lore, Stealth/Subterfuge. Used for non-magical pulls. [b]4. CUSTOM MAGECRAFT — Skill Design Worksheet[/b] Root Oracle doesn't ask you to pre-write individual spells. Instead, you design Skills: broad, versatile categories of magecraft that can produce a wide range of improvised effects at the moment you need them, not a fixed spell list. You still spend Prana and draw tokens exactly as before. What's changed is that the fiction of "what this can do" is a domain you interpret in the moment,. [b]4.1 The Skill Card[/b] Fill this out once per Skill, and keep it on a card: [hider=My Hider] Field- How to fill it in Name- What the mage calls this facet of their craft (e.g. "Reduction," "Coagulating Swordsmanship," "Kaleidoscope Thread") Foundation - Which of your Foundations (3.5) it draws on — sets the Rating you draw Domain - One or two sentences defining the conceptual boundary of what this Skill can touch — not a fixed effect. Broader domains are more versatile but harder to master convincingly; narrower domains are reliable but limited. ("Anything that can be stripped down to its most basic, essential form" is a Domain. "Shatter a lock" is not — that's a one-off trick, see 4.4.) Core Rank - 1–5. The Skill's baseline Difficulty when used squarely within its Domain. This is the number you actually reference at the table. Signature Applications - 3–5 example uses within the Domain, written as illustration, not an exhaustive menu. These exist so other writers immediately grasp the Skill's shape. Structural Drawback - A true, permanent limitation baked into how this Skill works — not chosen per-use. It grants a standing +1 token on every Central-use pull (4.2) This is optional [/hider] [b]4.2 Using a Skill: Scope Classification[/b] When a player wants their Skill to do something, state the intent in fiction, then classify how central that use is to the Skill's Domain. In solo/duo play the acting writer self-classifies honestly; in group scenes, any writer can propose a classification and the acting writer confirms, or the Ledger settles a genuine disagreement via an oracle question (1.1: "Is this within the Domain?"). Mystic Codes can be created using the same system, just have some common sense with the limitations and capabilities of such things. Prana Cost always equals the final adjusted Difficulty (before Elemental match discount, 3.3). Difficulty above your Circuit Quality (3.4) still costs 1 automatic Strain to force, same as any pull. [b]4.3 Worked Example[/b] Skill: "Reduction" — Foundation: Alchemy (4). Domain: anything that can be stripped down to its most essential, basic form — objects, arguments, wounds, even people's stated motives. Core Rank 3. Signature Applications: distill a poison to its purest toxin; strip a lie down to the bare fact underneath it; reduce a complex ward to its one load-bearing rune. Structural Drawback: the caster cannot build or complicate anything with this Skill, only simplify. Distilling a poison → Central, Difficulty 3, +1 token from the Drawback. Reducing a person's cover story to the one true motive underneath → Adjacent (people aren't quite "objects," but it's a natural reading), Difficulty 4. Trying to reduce a building to rubble in one shot → Peripheral (technically "reducing it to component parts," but it's straining the concept toward pure destruction), Difficulty 5, +1 Opposed added to the pull. [b] 4.4 Minor Utility Mysteries (one-off tricks)[/b] Not everything needs to be a Skill. For a single fixed, narrow trick that doesn't deserve its own Domain, things like General-Fundamentals that any trained mage picks up — just fill out a Name, Foundation, Effect (one sentence), Difficulty 1, Prana Cost 1, and skip 4.1–4.3 entirely. These carry no Structural Drawback and no Scope table. [b]5. THE BOUNTY BOARD — Proposing Missions[/b] There is no GM to hand out jobs, so the Organization runs on cases the players bring to the table themselves. Any player may post a Bounty: state in fiction who wants what done, for whom, and roughly why. That's the whole requirement. Whatever isn't already established stays genuinely unknown until someone asks. Is the client telling the truth? How dangerous is the target really? Is the pay fair? Don't decide these in advance— when it matters, ask the Oracle (1.1): state the question as a yes/no, judge its Likelihood from what's already true in the fiction, adjust the Ledger, and draw. The answer is real from that point on. • Any player may accept a posted Bounty for their character, alone or with others, and may pool staff or Prana with other department heads — agree how a payout splits before the job starts, not after. • If a player doubts a listing's accuracy, that's an Oracle question too (“Is the client's information accurate?”), asked before committing resources rather than assumed one way or the other. • Payout is agreed by ordinary consensus at posting — what feels fair for the risk and effort involved — the same way any other detail of the fiction gets settled at the table. If the table can't agree, ask the Oracle whether the client's offer is fair, generous, or lowball, and let that steer the number. • Completion follows directly from the mission's climactic pull (1.2): a clean Success pays in full, a complicated Success pays but with a string attached (a delay, a cut withheld, a favor instead of Marks), and a Failure pays nothing, though salvage or partial recovery is always fair game to argue for in fiction. A Root result is worth narrating as a windfall — extra pay, a rare item, a favor owed — at the table's discretion. [b]6. THE QUARTERMASTER'S LEDGER — Loot & Items[/b] There's no shop inventory to generate and no item tables to roll on. Marks earned from Bounties buy whatever it makes sense for the Organization, a contact, or a black market to actually have— common sense is the only rule. • If it's reasonable a place, person, or stash would have something, it's there. A quartermaster's cache has mundane gear and reagents; a black-market fence has stranger, pricier things; a target's safehouse has whatever that target plausibly needed. • When it isn't obvious whether something's available, or a player wants to search, loot, or scavenge for something specific, that's an Oracle question (1.1): “Does this place have a working Mystic Code?” “Is there anything worth salvaging here?” Judge Likelihood from the fiction and draw. • Prices work the same way payouts do (5.4) — agree a fair number for what it plausibly costs, or ask the Oracle if the asking price is fair, inflated, or a steal. • A Root result on a search or scavenging pull is the table's cue to make the find genuinely notable, something rarer or stranger than what was being looked for, not just more of the same. [/hider] If the CC and magecraft stuff is unwanted, could just use the oracle alone.