THE BOARD
IS IN.

Why nowSource changed after the last full integration. New focused tests and one runtime pass improved the evidence, but results have been mixed and the latest proof omits major game paths. Ownership, builds and proof do not currently name one stable candidate.

What to buildStop new mutations. Assign one owner and reconcile the protected project with the active work package. Record the exact source snapshot, current build, included changes and proof boundary. Compile after the final change, run the targeted streaming checks once, then run one full integration covering startup, traversal, harvesting and editing, inventory and UI, flight, save/load and representative six-biome behavior. Produce at most one current review build. Do not tune art, add systems, restart the accepted biome checkpoint, create more candidate builds or repeat failed verification without a concrete causal correction.

Player payoffBrian and future players receive one dependable build where improvements and defects belong to the same game state.

Done whenSource-present: one owner, frozen source snapshot, manifest and recovery point exist. Compiled: zero errors after the last source change. Tested: focused streaming checks and the current full integration pass. Runtime-proven: two consistent runs of the same build cover the named package paths without recurring allocator warnings, streaming-budget failures, missing floor or broken save/edit behavior. Brian-accepted: explicitly still pending after the technical candidate is ready.

Why nowThe opening teaches fourteen steps through Peggy and the ceramic forge but never reaches the pyre, appetite or Stone payoff. No current uninterrupted first-hour route or whole-package visual acceptance exists.

What to buildOn the frozen baseline, join arrival, starter gathering, Peggy's cave, the first worksite and one successful sacrifice into a single seed-driven route. Add only the hints, controller and focus behavior, accessibility cues, landmark visibility, placement clearance and reward feedback needed for that handoff. Reuse the existing pyre, appetite and Stone systems. Do not reopen the accepted biome checkpoint, hand-tailor a seed, add biomes or portals, start multiplayer or rewrite the whole tutorial.

Player payoffA new player learns what to do, why the world matters and what makes Worlds to Burn different before the opening loses momentum.

Done whenSource-present: the tutorial has an explicit first-sacrifice handoff and seed-driven route constraints. Compiled: the frozen candidate builds cleanly. Tested: sequence, failure recovery, keyboard/controller and deterministic route checks pass. Runtime-proven: a fresh save completes arrival through sacrifice without teleporting, dead travel, clipping, missing resources or unclear prompts. Brian-accepted: Brian approves the opening's clarity, pacing, fun and visual composition.

Why nowThe game says the Stone wants variety, but current appetite cross-recovery is not a chain mechanic. No explicit streak, escalation, break condition, reward ladder, feedback or persistence state was found.

What to buildAdd one small, explicit chain based on consecutive varied successful offerings. Define eligible variation, chain growth, break and reset behavior, visible preview and result feedback, and a bounded reward that feeds existing Stone or research progression. Persist it through the existing save model. Do not add a new currency, rewrite appetite, expand hidden rites, create a second pyre economy or rebalance every recipe.

Player payoffSacrifice becomes an understandable planning game: players choose what to burn next instead of repeating the single best offering.

Done whenSource-present: chain rules, state, feedback contract, reward and persistence exist separately from appetite recovery. Compiled: no errors on the frozen baseline. Tested: growth, invalid repetition, reset, reward bounds, save/load and appetite interaction pass. Runtime-proven: a normal sequence visibly builds, breaks and restores the chain correctly. Brian-accepted: Brian agrees the mechanic creates a clear, worthwhile choice without grind.

Why nowSix biome authorities and strong asset pipelines exist, but representative runtime proof is older and fragmented. Cave structure, placement coverage, water presentation, sky composition and creature habitat integration remain below whole-world acceptance.

What to buildExtend the composition and placement rules proven in the opening slice across the existing six biomes. Give each biome a bounded landmark hierarchy, cave threshold and interior identity, safe occupied footprints, creature grounding, water treatment and sky/atmosphere balance. Prioritize existing assets and migrate only the visible production set needed for representative routes. Do not add a seventh biome, authored water lists, fixed cave routes, seed-specific screenshot patches or another broad creature and model fleet.

Player payoffExploration produces recognizable places, readable routes, convincing habitats and memorable transitions instead of differently colored but similarly sparse terrain.

Done whenSource-present: reusable placement and composition rules cover the selected production set. Compiled: the frozen candidate builds cleanly. Tested: multi-seed placement, cave access, water support, unload/return and biome ownership checks pass. Runtime-proven: current ground and flight routes show all six identities without clipping, empty chambers, billboard-dominant skies or sheet-like water. Brian-accepted: Brian approves the six-biome world as cohesive and visually finished enough for the current milestone.

Why nowPyre, appetite, hidden rites, Stone powers, research, construction, automation and persistence are substantial in source, but no current normal-session proof shows them forming a stable long-term decision loop. Save performance and current-package coverage remain open.

What to buildAfter ranks one through three are stable, run a fresh normal session through gathering, construction, first sacrifice, Variety Chain, research choice, one automation payoff, one hidden-rite or Stone-power decision, save, exit, reload and continued play. Measure pacing, resource bottlenecks, save duration, restored state and player-facing feedback. Fix only defects or severe tuning blockers revealed by the session. Do not add endgame, new currencies, portal progression, multiplayer or a broad economy redesign.

Player payoffThe game continues to offer understandable choices after onboarding, and the player can safely return to a world without losing progress or strategy.

Done whenSource-present: all exercised systems and persistence fields exist on the frozen baseline. Compiled: the final session candidate builds cleanly. Tested: focused economy and save/load regressions pass. Runtime-proven: one uninterrupted 30–60 minute session plus reload preserves world edits, machines, research, ritual state, appetite, Variety Chain and rewards within acceptable timing. Brian-accepted: Brian agrees the session stays interesting, readable and trustworthy.