v0.4.0 vNext — closed-loop world model + settlement contract (#29)

* feat(v0.4.0): vNext — closed-loop world model + settlement contract

Implements the vNext architecture from the research doc: demote the noisy
multi-rail planner in favour of a closed loop (world truth → invariant check)
plus a single utility-driven goal authority.

L1 services (fix no_drop / silent pathfinder hang first):
- InventoryLedger: diff-based "did I actually get it" verifier; acquire-food
  now confirms via ledger.gainedSince instead of the unreliable count/event.
- MotionService.gotoSafe: wall-clock timeout + progress watchdog +
  path_update(noPath/timeout) → structured {reached|stuck|timeout|nopath}.

L3 plan — unify the three competing rails (curriculum/manifesto/storyline):
- Settlement Contract: ordered M0–M9 milestones, each invariant-checked
  against an authoritative world view (early steps delegate to the proven
  curriculum; late game adds farming).
- InvariantChecker + predicate library; GoalManager selects the lowest unmet
  milestone via utility argmax (food-urgency preempts, DEPS-style).
- Wired into the scheduler: bot.js precomputes snapshot.contract; reflex.js
  consumes it in place of the storyline rail. Manifesto L0 still preempts.

Eval + robustness:
- Village Score (single 0..1 metric) on the snapshot + TUI "build" line.
- survive.dig-in skill + dusk_dig_in mode (exposed at night, no bed → cover).
- approach_block helper (GoalNear + lookAt, avoids GoalLookAtBlock #341).

+28 new tests (450 total green). LLM remains entirely off the tick path.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* feat(v0.4.0): finish vNext plan — anti-loop, skill-graph, worldDelta diff, flee→motion

Completes the remaining v0.4.0 plan items and one fix motivated by a live
in-game observation (flee hanging 30s against a persistent zombie).

- flee → MotionService.gotoSafe: structured {stuck|timeout|nopath} in ~4s with
  a blind-retreat fallback, instead of the observed 30s pathfinder hang + 3
  watchdog replans. Movements setup guarded so it is unit-testable.
- QW5 anti-loop (runtime/anti-loop.js): same skill failing >=3x in 5min →
  30min blacklist (reflex shouldSkip) + one-shot improvement_request
  (bot.js drainFired -> writeProposal).
- 4.1 closed-loop worldDelta: runSkill snapshots inventory before execute and
  attaches the real delta (_invObserved) to every successful result; opt-in
  skill.expectGain asserts the claimed gain or returns world_unchanged.
- 3.6 skill-graph (Plan4MC): declarative requires/produces for ~20 skills;
  prerequisitesMet/canRun/runnableFrontier; GoalManager annotates suggestions
  with blockedBy when prereqs are unmet.

+22 tests (472 total green). Live smoke confirmed dig-in works and no new
errors; flee loop is what this commit's flee migration addresses.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

---------

Co-authored-by: Yuriy Mayatnikov <mayatnikov@me.com>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>

runtime/reflex.js

@@ -528,6 +528,18 @@ function curriculumReflex(ctx) {
 	const storyStep = ctx.disableStoryline ? null : pickCurrentStep(s);
 	if (storyStep) ctx.storyStep = storyStep;
 	const storySkillId = (storyStep && !storyStep.emergency && storyStep.suggestion?.skillId) ? storyStep.suggestion.skillId : null;
+
+	// v0.4.0 — Settlement Contract is the unified progression authority,
+	// replacing the storyline rail (which competed with the manifesto). It is
+	// precomputed in bot.js (snapshot.contract) via the GoalManager: lowest
+	// unmet milestone, with food-urgency utility preemption. Manifesto L0
+	// (alive emergencies) still preempts it; storyline/curriculum remain as
+	// fallbacks when the contract is disabled (tests) or has no suggestion.
+	const contractGoal = ctx.disableContract ? null : (s.contract ?? null);
+	if (contractGoal) ctx.contractGoal = contractGoal;
+	const contractSkillId = (contractGoal && !contractGoal.done && contractGoal.suggestedSkill?.skillId)
+		? contractGoal.suggestedSkill.skillId
+		: null;
 	const metricRecovery = metricRecoverySkill(ctx, plan?.skillId);
 	if (metricRecovery) {
 		ctx.lastCurriculumAt = Date.now();
@@ -584,7 +596,7 @@ function curriculumReflex(ctx) {
 	// First hint → small wander (might just be 32-block reach issue).
 	// Every subsequent hint while still inside the backoff window → use
 	// explore.far so the bot actually leaves the patch it's stuck in.
-	if ((!plan?.skillId && !manifestoSkillId && !storySkillId) || wantWander) {
+	if ((!plan?.skillId && !manifestoSkillId && !storySkillId && !contractSkillId) || wantWander) {
 		ctx.lastCurriculumAt = Date.now();
 		const fallbackId = wantWander && consecutiveWanderHints >= 1 ? "explore.far" : "wander";
 		// v0.2.0-rc.3 — consult advice on the FALLBACK dispatch too. Without
@@ -633,13 +645,16 @@ function curriculumReflex(ctx) {
 	const manifestoEmergency = activeNeed?.need?.level === 0;
 	let skillId, skillSource;
 	if (manifestoEmergency) {
-		skillId = manifestoSkillId ?? storySkillId ?? plan.skillId;
+		skillId = manifestoSkillId ?? contractSkillId ?? storySkillId ?? plan?.skillId;
 		skillSource = `manifesto:${activeNeed.need.id}`;
+	} else if (contractSkillId) {
+		skillId = contractSkillId;
+		skillSource = `contract:${contractGoal.milestone.id}`;
 	} else if (storySkillId) {
 		skillId = storySkillId;
 		skillSource = `storyline:${storyStep.step.id}`;
 	} else {
-		skillId = manifestoSkillId ?? plan.skillId;
+		skillId = manifestoSkillId ?? plan?.skillId;
 		skillSource = manifestoSkillId ? `manifesto:${activeNeed.need.id}` : "curriculum";
 	}
 
@@ -697,6 +712,14 @@ function curriculumReflex(ctx) {
 		}
 	}
 
+	// QW5 anti-loop: this skill failed ≥3× in 5 min → it's blacklisted. Skip
+	// and nudge toward exploration so we leave the situation that loops it.
+	if (ctx.antiLoop?.shouldSkip(skillId)) {
+		ctx.skillBackoff = ctx.skillBackoff ?? {};
+		ctx.skillBackoff["__wander_hint__"] = Date.now() + SKILL_BACKOFF_MS;
+		return { action: "noop", kind: "anti-loop-blacklisted", label: skillId };
+	}
+
 	// v0.2.0 — consult learned lessons. If a high-confidence lesson says
 	// "avoid <skillId> in this situation", swap to its preferred
 	// alternative (or back off entirely if no safe alternative is named).
@@ -719,6 +742,9 @@ function curriculumReflex(ctx) {
 	ctx.dispatch(() => runSkill(dispatchSkillId, ctx, dispatchArgs), dispatchSkillId, {
 		onComplete: (res) => {
 			ctx.skillBackoff = ctx.skillBackoff ?? {};
+			// QW5 anti-loop bookkeeping: feed every outcome so repeated failures
+			// of the same skill get detected, blacklisted and ticketed.
+			ctx.antiLoop?.record({ skillId: dispatchSkillId, ok: !!res?.ok, code: res?.code ?? null });
 			if (advice.lessonId) reportAdviceOutcome({ lessonId: advice.lessonId, succeeded: !!res?.ok });
 			if (appliedRecommendationId) {
 				markRecommendationOutcome(appliedRecommendationId, {