Shroom and Gloom asks you to build two connected decks. Combat cards solve fights; Explore cards navigate and reshape the run. Read each offer for its immediate job, its setup cost, and the payoff it can reach with cards you already own.
Combat and Explore cards solve different problems.
Steam's current description confirms that every run develops a Combat deck and an Explore deck. Combat cards handle enemies. Explore cards move through the dungeon and can modify cards, unlock weapons, or open abilities. Cards from both decks can grow and change during a run.
Ask how a card helps you deal damage, control a threat, survive pressure, or enable an existing payoff.
Ask what the card changes outside combat: movement, access, card modification, recovery, or future options.
An Explore choice can strengthen the next fight, while combat losses can change what your route must provide.
Steam says each current pathway contains roughly 300 Explore and Combat cards. That is a per-pathway marketing estimate, not proof of 600 unique, mutually exclusive cards.
Read a card in layers.
Before calling a card strong or weak, separate what is printed from what the run supplies. Exact terms and numbers should be checked against the current in-game card, because Early Access patches can change them.
- Deck and job: Is it a Combat or Explore card, and what useful action does it perform immediately?
- Cost: What energy, resource, card state, or board condition must be present?
- Target: Does the effect apply to this card, another card, an enemy, or the wider encounter?
- Timing: Is it played directly, triggered by another event, or exposed through an exploration interaction?
- Scaling: Which value can grow, and what does the run need to supply for that growth?
- Persistence: Is the change temporary, limited-use, consumed, or retained? If the current UI does not make that clear, treat it as unverified.
Do not flatten every effect into one line.
A build-scoped, unofficial index for public build 25221077 separates ordered play actions, exploration interaction actions, and triggered actions. That distinction is useful: a card's normal play effect may not describe everything it can do when modified, interacted with, or triggered by another system.
| Layer | What to check | Why it matters |
|---|---|---|
| Played action | The ordered effect when the card is used | Order can change whether setup arrives before payoff. |
| Explore interaction | Choices or modifications offered outside combat | The same card can have a different role while navigating. |
| Triggered action | The event that causes the effect to fire | A powerful line may be dormant without its trigger. |
| Variant | Upgrade, interaction, or development version | Matching names do not guarantee matching behavior. |
Use a four-question offer check.
- What problem does this solve now? Name a fight, route, recovery, or consistency problem rather than a theme.
- What enables it? Count the cards and resources that already create the required condition.
- What pays it off? If the answer depends on a card you hope to find later, price that uncertainty into the choice.
- What draw does it replace? A thematic card can still dilute a working deck when drawn without its partner.
This framework is editorial guidance, not a hidden scoring formula. It is meant to keep a run functional while cards grow and synergies emerge.
Current videos show outcomes, not universal recipes.
A September 12 Dabble video is titled around an infinite deck and a one-shot final boss. It confirms that players are exploring loops and burst finishes in the current Early Access window, but the title alone does not establish the full card list, minimum components, substitutions, or failure cases.
Use this as a case study for setup and payoff, not proof that every run can reproduce the result.
A September 10 Janitor/Facility run from Zueljin Plays shows Data, Machines, Gunk, deck thinning, and a late build pivot. That run ends under final-boss discard pressure. Together, the examples show why a build guide must document setup and disruption, not merely the winning interaction.
Watch the infinite-deck case in the player above, then compare its setup with the embedded Janitor run without leaving this guide site.
A database record is not an obtainable-card promise.
The unofficial index exposes 907 card pages, but its own accounting distinguishes 900 pages associated with the current build from seven retained pages no longer present in that build. Records may also include variants or content not proven to appear in a normal run.
This site therefore does not claim that 907 cards are obtainable. A future individual card entry needs its deck, class, energy, displayed effect, interaction or trigger, build, source, and current availability checked before exact values are published.
- Gunk, Warmth, Uses, Temporary, Consume, and Exhaust still need current-UI verification before definitive glossary entries.
- Bless and resting need their current targets and persistence checked.
- “Mutated cards” may be a formal mechanic, an update label, or creator wording; it remains unverified.
- Demo build 23718635 values do not substitute for Early Access build 25221077.
Sources and version notes.
Official current overview: Steam store, checked September 13, 2026, for the two-deck structure, card growth, food interactions, and the approximate per-path card count.
Build metadata: SteamDB public branch, showing build 25221077 when checked September 13, 2026. SteamDB is used for build identity, not gameplay rules.
Unofficial build index: Cards, mechanics, and glossary. These are build-scoped research aids and not an official or guaranteed drop list.
Player examples: Dabble and Zueljin Plays videos published after the September 10 Early Access launch. Each represents one run.