Currency & Economy Model — Credits, Reputation, Mastery, Knowledge

Companion docs:

mission-objectives.md — the goal/reward layer: a goal’s :reward cells are denominated in these currencies. That doc owns per-goal rewards; this doc owns the currencies themselves and the sinks.
mission-control.md — contract generation, threat tier, the board
../cartridges/authoring/campaign-economy.md — payout formula, self-correcting economy, tool-pricing standardization (the credit income side)
deck-state.md — UDS struct (credit_balance, reputation, cartridge_history); gains a per-cart mastery counter + dossier-key store (§9)
../cartridges/modules/the-vault.md — knowledge_index + DOSSIER, the knowledge currency’s home Related ADRs: ADR-0043 (objective model — the reward layer), ADR-0030 (threat-cap inheritance — mastery climbs toward it), ADR-0042 (first-party programs — the credit upgrade tree), ADR-0028 (Mission Control), ADR-0011 (UDS persistence). An ADR is owed (§9) — it will amend ADR-0030 (mastery → threat range) and ADR-0042 (program-upgrade tiers).

1. Scope

The objective model (mission-objectives.md) made a goal’s reward first-class but deferred “what each currency is for” to tuning. This doc answers that: it defines the four currencies, the guardrail that keeps them distinct, and the credit sinks that give the spend currency somewhere to go. It is the economy beneath the reward field — mission-objectives.md says a goal pays (rep 4 :on-resolve); this doc says what reputation is and does.

Two decisions anchor it (2026-06-21 brainstorm):

Four currencies with distinct feels (§2) — a currency with no felt value is worse than none.
A spend guardrail (§3) — credits can’t buy the other three.

2. The four currencies

Currency	Feel	Earned by	Volatility	What it does
Credits `¤`	wallet	mission `¤` payouts	spend-down	rig/program upgrades, buy access, consumables (§5)
Reputation	résumé	mission success	can drop (fail / abandon)	global baseline threat tier + better jobs (§6)
Mastery (per-cart)	muscle memory	threat-weighted contracts cleared with that cart	monotonic, no decay	unlocks that domain’s threat range (→ its `:threat-cap`) + a spices ladder (§7)
Knowledge	library	recon / Vault / intel goals	accumulates (in DOSSIER)	typed keys hard-gate specific missions + soft-odds bias (§8)

The reason the set works: credits flow out; reputation, mastery, and knowledge accumulate. Wallet, résumé, muscle memory, library — four different relationships with the operator, so a goal that pays reputation instead of credits genuinely feels different, not like a smaller bounty. And because the three accumulating currencies don’t self-correct the way ¤ does (campaign-economy.md), the late game naturally shifts onto them.

3. The guardrail

Credits buy shortcuts and gear. They never buy capability or standing.

A fat wallet can skip a step or upgrade your kit — but it cannot buy mastery, cannot buy knowledge, cannot buy reputation. Doing the work is the only way to get better. This one rule protects the three accumulating currencies from being trivialized by a rich operator, and it is the boundary that makes “buy access” (§5) a shortcut rather than a win button: you bought your way in; you didn’t get better, and you earned less of the currencies you’d have earned by doing it yourself.

4. Where each currency attaches

The currencies show up at three different layers — keeping them straight is what keeps the model legible:

Layer	Mechanism	Currencies
Goal reward (one objective pays)	a `:reward` cell in `mission-objectives.md`	`¤` · `rep` · `intel` · `access`
Completion accrual (finishing the contract grants)	computed at `complete-mission`	`rep` delta (canon) + `mastery` (per-cart, threat-weighted)
Prerequisite / unlock (gates what you can do)	checked at accept / phase / completion	knowledge keys (hard-gate) · access (bought or granted)

Notes that fall out of this:

Mastery is not a goal reward — you earn it by clearing the whole contract, not by ticking a box. (Keeps it from being farmed off trivial objectives.)
access is not a banked currency — it’s a binary unlock, either bought with ¤ (§5) or granted by a goal :reward. It does not accumulate.
Knowledge does double duty — a goal reward (intel, which deposits a key) and a mission prerequisite (§8).

5. Credits & the sinks (hybrid: tree + churn)

Credits are the only spend-down currency, so they need real sinks or they pile up and die. The model is hybrid: a permanent progression tree plus ongoing churn. Three buckets:

5.1 Rig upgrades — the tree (permanent progression)

The upgrade tree is the ADR-0042 program roster. “Upgrade software on the rig” means buying better tiers of the thirteen first-party programs:

Harden knSALK (survive counter-intrusion) · stealthier CONDUIT sessions · deeper bzbx scans · bigger DOSSIER storage · faster Kommander transfers · harder cracks in Keyring.
Plus cross-cutting rig stats: more Lambda slots, faster trace decay (a “stealth rig” build that directly eases ICE BREAKER play).

Permanent, accumulative — this is the operator’s build. A program tier purchased never expires.

5.2 Access / shortcuts — the cross-currency exchange (situational)

Spend ¤ to skip work: buy a backdoor (skip the breach, start deeper), buy a dossier key outright (¤ → knowledge), buy a fixer tip (reveal a latent goal or surface an opportunity contract). Per the guardrail (§3), bought access earns less mastery/knowledge than doing it yourself — you got through, you didn’t learn. This is the deliberate pay-to-skip-recon lever, and the only place ¤ touches the other currencies (and only by substituting for them, never buying them).

5.3 Consumables / panic gear — the churn (per-run attrition)

Extra tool charges, a one-shot trace-wipe or HUNTER-stall, an emergency extract, failure insurance. Per-run, consumed on use — the steady drain that keeps credits flowing out even after the tree is bought.

Why all three: bucket 1 is progression, bucket 3 is attrition, bucket 2 is the exchange. Once the tree is bought out, the self-correcting economy (campaign-economy.md) compresses income while consumables + prestige sinks mop up surplus, so ¤ stays meaningful long-term.

6. Reputation — global standing

Reputation is canon (mission-control.md §3.1, deck-state.md): a single global number that sets the operator’s baseline threat tier — how hard the jobs the board offers can be — and gates better contracts. It is the volatile currency: it rises on success and drops on failure/abandonment (mission-control.md §5.4). It is your standing in the world — what gets you in the door — distinct from your fluency with a tool (§7).

7. Mastery — per-cart fluency

New in this model. Mastery is a per-cartridge counter that advances on contracts cleared with that cart, threat-weighted (harder jobs teach you more), monotonic, no decay — you don’t forget a tool.

What mastery unlocks, within that cart’s domain:

Threat range (mechanical). Global reputation (§6) sets your baseline tier; mastery lifts the reachable threat band inside a domain toward that cart’s :threat-cap (ADR-0030). ICE BREAKER’s cap is 6; a low-mastery operator only sees its threat 1–3 contracts even at decent global rep, and mastering it opens the 4–6 band for that domain specifically.
Spices (content). Each cartridge ships a mastery ladder (its author’s job) mapping thresholds to: new tools / verbs / CONS combos, new mission types in the domain, and rare targets / named adversaries.

This yields a real generalist vs. specialist axis: broad global rep = harder jobs across the board; deep mastery = you punch above your rep inside your chosen domains. A mid-rep ICE BREAKER specialist can run threat-6 netruns while still being middling at forensics.

Mastery is its own subsystem (§9): it needs a per-cart counter in deck state, a per-cart :mastery-ladder declaration in cart registration, and it amends the ADR-0030 threat gate. It is specced here at the model level; the build spec + ADR are follow-ons.

8. Knowledge — typed keys + the dossier

Knowledge is the library currency, living in DOSSIER / the Vault (the-vault.md). It works at two strengths:

Hard gate — typed keys (specific). Knowledge is discrete dossier keys (RELAY-7-TOPOLOGY, TANAKA-CREDENTIALS, BANK-ROUTING). Some missions cannot be finished without the right keys — a goal or phase declares the keys it requires, and without them the path is blocked. This is what makes recon specifically load-bearing: you case this target, the DOSSIER becomes a real inventory, and “buy a dossier key” (§5.2) is a literal substitution. (Keys are deposited by intel goal rewards and by recon/Vault play.)
Soft bias — the index score. The existing knowledge_index (the-vault.md, payout-bias) remains as a breadth layer: general expertise that improves odds / payouts / domain access, separate from the specific-key hard gate.

The loop this creates: recon/Vault → keys in DOSSIER → gated missions become finishable → completion → more credits/rep/mastery → harder targets that need more recon. Recon stops being flavor and becomes a prerequisite tier of play — the realization of the original “some missions require recon” intent.

9. Open areas (design continuing)

Magnitudes / balancing — relative value of ¤ ↔ rep ↔ mastery ↔ a key; sink pricing (program tiers, access, consumables); the threat-weighting curve for mastery gain.
Per-cart mastery ladders — the actual spice content per cartridge (ICE BREAKER’s 7th tool, BLACK LEDGER’s advanced audits, …) — authored per cart.
Deck-state fields — the per-cart mastery counter and the dossier-key store (deck-state.md / UDS). Embedded owns the byte layout.
The ADR — owed once the model firms; it will amend ADR-0030 (mastery → threat-range climb) and ADR-0042 (program-upgrade tiers), and reconcile with the campaign-economy.md payout formula.
Pay-to-skip economics — exact mastery/knowledge penalty for bought access vs. earned (§5.2); pricing access against the recon it replaces.

10. Relationship to existing docs

mission-objectives.md owns the per-goal reward (:reward cells). This doc owns the currencies those cells are denominated in, plus the sinks. No duplication: that doc references this for “what a currency is,” this references that for “how a goal pays.”
campaign-economy.md owns credit income (the payout formula, self-correction, tool pricing). This doc owns credit outflow (the sinks). The two meet at the ¤ balance.
ADR-0030 owns the threat-cap; §7 here climbs toward it via mastery (an amendment, owed).
ADR-0042 owns the program roster; §5.1 here makes those programs the upgrade tree (an amendment, owed).