TestGame - Development Epics

These epics sequence complete player-visible capabilities. They describe design outcomes, not engine architecture or implementation classes.

Epic 1: Play Space and Interaction

Outcome: The player can enter the compact sandbox, understand its zones, navigate the camera, select actors, issue commands, and place every required Building.

High-Level Stories

1. Present the production, Deer, and two Monster zones with readable world labels.
2. Pan, rotate, and zoom the camera within bounds that preserve zone readability.
3. Select one or multiple units and issue contextual move or attack commands.
4. Create and recall control groups.
5. Preview and place each available Building on a valid pad with explicit invalid-state feedback.
6. Keep Storage, selection, active Quests, and job timers visible across interactions.

Completion Evidence

• Every required Building can be placed without blocking unit routes.
• Invalid placement changes no state and explains why it failed.
• All five player units can be selected and commanded using the documented input model.

Epic 2: Mythical Storage and Production

Outcome: Raw Resources enter shared Storage and become intermediate materials through visible, cancellable, concurrent jobs.

High-Level Stories

1. Begin each session with the documented starting Resource quantities.
2. Generate each configured raw Resource at 1 per 10 seconds.
3. Display exact Storage quantities, recent changes, and each Resource Building’s active generation cadence.
4. Start a valid 30-second Processing job and commit its inputs.
5. Run one job per Building while allowing separate Buildings to work concurrently.
6. Keep each active job’s output, progress percentage, and remaining time visible after its Building panel closes.
7. Cancel a job for a full refund without duplicated or lost Resources.
8. Deliver every intermediate output to Mythical Storage.

Completion Evidence

• Every intermediate in the canonical production table can be produced.
• Concurrent jobs maintain independent timers and exact Resource accounting.
• Start, completion, cancellation, and blocking states are visible and numerically correct.
• Active Processing and Crafting interfaces provide a visible Cancel action; a busy Building names that blocker.

Epic 3: Crafting, Inventory, and Equipment

Outcome: The player can turn intermediates into Crafted Goods, receive them in reserved slots, and apply their exact effects to eligible units.

High-Level Stories

1. Browse canonical Recipes with exact inputs, availability, outputs, and duration.
2. Reserve one of ten Inventory slots before committing Craft inputs.
3. Block Craft start when an input or slot is missing and state the remedy.
4. Complete a Craft into its reserved slot, or cancel it with a full input refund and slot release.
5. Distinguish empty, reserved, occupied, focused, equipped, and ineligible slots.
6. Equip an Iron Sword to an eligible melee unit and show the +5% HP change.
7. Upgrade T1 Leather Armor into T2 Metal Armor and equip it to a Soldier for +10% HP.
8. Preserve correct current/max HP when equipment is added or removed.

Completion Evidence

• All four Crafted Goods can be produced through their full chains.
• Full Inventory and missing-input cases never consume Resources.
• Before/after unit values make the sword and armor effects unambiguous.
• Craft completion names the Crafted Good and visibly changes its reserved slot to occupied.

Epic 4: Units, Combat, and Morale

Outcome: The starting force can fight two Monster groups, hunt Deer, and visibly benefit from permanent and temporary crafted effects.

High-Level Stories

1. Spawn the documented player roster, Monsters, and Deer with initial tuning values.
2. Move, acquire targets by command, attack automatically, take deterministic damage, and resolve defeat.
3. Recover defeated player units at the starting area after 10 seconds without losing session progression.
4. Run simple bounded Monster acquisition, pursuit, and return behavior.
5. Make Deer flee and transfer 1 Leather to Storage on defeat.
6. Deploy a Morale Flag within range of a Soldier.
7. Show its 10 m area, affected Soldiers, +5% HP effect, and 10-second timer.
8. Remove the Flag and effect cleanly without stacking or salvage.
9. Expose HP, damage values, damage source, and active effects throughout combat.

Completion Evidence

• Both encounter groups can be engaged and defeated; move/attack commands either execute or show a documented invalid-state response, and no unit remains indefinitely stuck in acquire, pursue, or return.
• Leather acquisition supports the armor and Morale chains.
• Equipment and Flag effects can be compared using visible exact values.

Epic 5: Loot, Hero Skills, and Quests

Outcome: Monster loot expands Hero actions and both strategy paths produce visible Quest progress and acknowledgment.

High-Level Stories

1. Drop one Slash Scroll and one Fireball Scroll from designated Monsters.
2. Distinguish dropped, learned, and equipped skill states.
3. Teach and equip each Scroll only to its eligible Hero.
4. Target, preview, execute, and cool down Slash and Fireball using documented values.
5. Make both Quests available from session start and accept either independently.
6. Initialize accepted Quest progress from session counters so earlier qualifying events are preserved.
7. Track the accepted Aggressive Quest from 0/10 to 10/10 Monster defeats.
8. Track the accepted Production Quest from incomplete to one completed T2 Metal Armor.
9. Acknowledge completion with a transient popup and persistent completed tracker state, without material reward.

Completion Evidence

• Dropped, learned, and equipped states are visibly distinct; invalid targets consume neither cooldown nor skill state.
• Dropped, learned, and equipped Skill states are visually distinguishable.
• Quest counts update once per qualifying event and cannot duplicate completion.
• Both Quests remain concurrently visible and progress independently; completing either leaves the other active and unchanged except for shared qualifying events.

Epic 6: UX, Accessibility, and Feedback

Outcome: The full experiment is understandable from the game surface and remains operable at its supported desktop layouts.

High-Level Stories

1. Apply the restrained command-table visual hierarchy and consistent state colors.
2. Keep Storage, selection context, active Quests, and active timers persistent.
3. Present exact blockers and remedies without mutating state.
4. Provide visible keyboard focus for every keyboard-operable screen control.
5. Scale the UI with supported window changes without losing critical information.
6. Communicate every state with text or iconography in addition to color.
7. Keep nonessential motion restrained while preserving timing and effect legibility.
8. Anchor placement validity, dropped Scrolls, affected Soldiers, and damage values to their world subjects.

Completion Evidence

• The production, crafting/equipment, combat/loot, and Quest journeys can each be completed without hover-only information.
• Required persistent information remains available after resize and scaling.
• A silent run retains full causal readability.

Epic 7: Baseline Evaluation and Stabilization

Outcome: Tim can repeatedly complete and judge the whole experiment using evidence rather than implementation completion alone.

High-Level Stories

1. Complete an aggression-first pass and record perceived payoff and friction.
2. Complete a production-first pass and record perceived payoff and friction.
3. Explain the exact sword, armor, and Flag effects from the game surface.
4. Complete three consecutive end-to-end sessions without inspection or state mutation through Unity or development tools.
5. Run the ten-minute performance scenario on Tim’s reference PC and record average FPS, 1% low, and the reference configuration.
6. Record every tuning change with its value and reason; retest core-journey changes.
7. Make and record a continue, extend, rethink, or stop decision.
8. Capture the evaluation using the PRD’s evaluation-template.md prompts.
9. If continuing, define the Extension Milestone’s hypothesis, smallest scope, success criteria, counter-metrics, and relationship to Baseline before expanding scope.

Completion Evidence

• Technical and gameplay success metrics in gdd.md have recorded pass/fail results and observations.
• Known failures and scope pressures are recorded rather than hidden.
• Any proposed extension is separable from the completed Baseline.

Epic Sequence and Dependencies

1. Epic 1 establishes the interaction surface required by every later capability.
2. Epic 2 supplies authoritative Resource and job behavior.
3. Epic 3 establishes crafting, Inventory, eligibility, and equipment state; its tactical-effect evidence completes with Epic 4’s unit health model.
4. Epic 4 establishes units and combat, then completes the observable equipment and Morale effects supplied by Epic 3.
5. Epic 5 depends on combat, crafting completion, and visible event feedback.
6. Epic 6 begins alongside Epic 1, supplies the shared feedback and accessibility contract across Epics 1–5, and receives final end-to-end validation after Epic 5.
7. Epic 7 begins only when the full Baseline journey is playable.