@10et/cli 1.0.0

package/scripts/test-map-eventbus.sh

@@ -0,0 +1,357 @@
+#!/usr/bin/env bash
+# MAP Event Bus + Peter Parker — Integration Test
+#
+# Run from the jfl-cli directory:
+#   ./scripts/test-map-eventbus.sh
+#
+# What this tests:
+#   1. Context Hub starts with event bus
+#   2. Publish events via REST
+#   3. Retrieve events with pattern filtering
+#   4. SSE real-time streaming
+#   5. WebSocket streaming (if wscat installed)
+#   6. Peter Parker config generation (all 3 profiles)
+#   7. Peter Parker status reads from live bus
+#   8. MCP tool simulation (events_publish, events_recent)
+#   9. Event persistence across restart
+#   10. Pi CLI detection check
+
+set -euo pipefail
+
+RED='\033[0;31m'
+GREEN='\033[0;32m'
+YELLOW='\033[1;33m'
+CYAN='\033[0;36m'
+GRAY='\033[0;90m'
+NC='\033[0m'
+
+PASS=0
+FAIL=0
+SKIP=0
+
+pass() { ((PASS++)); echo -e "  ${GREEN}✓${NC} $1"; }
+fail() { ((FAIL++)); echo -e "  ${RED}✗${NC} $1"; }
+skip() { ((SKIP++)); echo -e "  ${YELLOW}→${NC} $1 (skipped)"; }
+
+PROJECT_ROOT="$(cd "$(dirname "$0")/.." && pwd)"
+cd "$PROJECT_ROOT"
+
+echo -e "\n${CYAN}═══════════════════════════════════════════════════════${NC}"
+echo -e "${CYAN}  MAP Event Bus + Peter Parker — Integration Tests${NC}"
+echo -e "${CYAN}═══════════════════════════════════════════════════════${NC}\n"
+
+# ─── 0. Build check ──────────────────────────────────────────────────
+echo -e "${YELLOW}[0] Build${NC}"
+if [ -f dist/index.js ] && [ -f dist/lib/map-event-bus.js ]; then
+  pass "dist/ exists with MAP modules"
+else
+  echo -e "${RED}Build missing. Run: npm run build${NC}"
+  exit 1
+fi
+
+# ─── 1. Start Context Hub ────────────────────────────────────────────
+echo -e "\n${YELLOW}[1] Context Hub Startup${NC}"
+node dist/index.js context-hub stop 2>/dev/null || true
+sleep 1
+OUTPUT=$(node dist/index.js context-hub start 2>&1)
+if echo "$OUTPUT" | grep -q "started"; then
+  pass "Context Hub started"
+else
+  fail "Context Hub failed to start"
+  echo "$OUTPUT"
+  exit 1
+fi
+
+sleep 1
+
+TOKEN=$(cat .jfl/context-hub.token 2>/dev/null)
+if [ -z "$TOKEN" ]; then
+  fail "No auth token found"
+  exit 1
+fi
+pass "Auth token exists"
+
+# Find port from health check
+PORT=""
+for p in 4521 4200 4201 4300; do
+  if curl -s "http://localhost:$p/health" 2>/dev/null | grep -q "ok"; then
+    PORT=$p
+    break
+  fi
+done
+
+if [ -z "$PORT" ]; then
+  fail "Cannot find Context Hub port"
+  exit 1
+fi
+pass "Context Hub responding on port $PORT"
+
+BASE="http://localhost:$PORT"
+AUTH="Authorization: Bearer $TOKEN"
+
+# ─── 2. Publish Events ───────────────────────────────────────────────
+echo -e "\n${YELLOW}[2] Publish Events${NC}"
+
+# Publish various event types
+RESP=$(curl -s -X POST "$BASE/api/events" \
+  -H "$AUTH" -H "Content-Type: application/json" \
+  -d '{"type":"session:started","source":"test-runner","data":{"branch":"test"}}')
+if echo "$RESP" | grep -q '"id"'; then
+  pass "session:started event published"
+else
+  fail "Failed to publish session:started"
+fi
+
+RESP=$(curl -s -X POST "$BASE/api/events" \
+  -H "$AUTH" -H "Content-Type: application/json" \
+  -d '{"type":"task:completed","source":"test-runner","data":{"task":"build","duration":12}}')
+if echo "$RESP" | grep -q '"id"'; then
+  pass "task:completed event published"
+else
+  fail "Failed to publish task:completed"
+fi
+
+RESP=$(curl -s -X POST "$BASE/api/events" \
+  -H "$AUTH" -H "Content-Type: application/json" \
+  -d '{"type":"peter:started","source":"peter-parker","data":{"profile":"balanced"}}')
+if echo "$RESP" | grep -q '"id"'; then
+  pass "peter:started event published"
+else
+  fail "Failed to publish peter:started"
+fi
+
+RESP=$(curl -s -X POST "$BASE/api/events" \
+  -H "$AUTH" -H "Content-Type: application/json" \
+  -d '{"type":"decision:made","source":"claude-code","data":{"decision":"use-postgres","reason":"cost"}}')
+if echo "$RESP" | grep -q '"id"'; then
+  pass "decision:made event published"
+else
+  fail "Failed to publish decision:made"
+fi
+
+# Test validation
+RESP=$(curl -s -X POST "$BASE/api/events" \
+  -H "$AUTH" -H "Content-Type: application/json" \
+  -d '{"data":{"bad":"no type or source"}}')
+if echo "$RESP" | grep -q '"error"'; then
+  pass "Rejects event without type/source"
+else
+  fail "Should reject event without type/source"
+fi
+
+# Test no auth
+RESP=$(curl -s -X POST "$BASE/api/events" \
+  -H "Content-Type: application/json" \
+  -d '{"type":"custom","source":"x","data":{}}')
+if echo "$RESP" | grep -q "Unauthorized"; then
+  pass "Rejects unauthenticated request"
+else
+  fail "Should reject unauthenticated request"
+fi
+
+# ─── 3. Retrieve Events ──────────────────────────────────────────────
+echo -e "\n${YELLOW}[3] Retrieve & Filter Events${NC}"
+
+RESP=$(curl -s "$BASE/api/events?limit=10" -H "$AUTH")
+COUNT=$(echo "$RESP" | python3 -c "import sys,json; print(json.load(sys.stdin)['count'])" 2>/dev/null || echo 0)
+if [ "$COUNT" -ge 4 ]; then
+  pass "Retrieved $COUNT events (expected >= 4)"
+else
+  fail "Expected >= 4 events, got $COUNT"
+fi
+
+# Pattern filter
+RESP=$(curl -s "$BASE/api/events?pattern=peter:*&limit=10" -H "$AUTH")
+COUNT=$(echo "$RESP" | python3 -c "import sys,json; print(json.load(sys.stdin)['count'])" 2>/dev/null || echo 0)
+if [ "$COUNT" -ge 1 ]; then
+  pass "Pattern filter peter:* returned $COUNT event(s)"
+else
+  fail "Pattern filter peter:* should return >= 1"
+fi
+
+RESP=$(curl -s "$BASE/api/events?pattern=session:*&limit=10" -H "$AUTH")
+COUNT=$(echo "$RESP" | python3 -c "import sys,json; print(json.load(sys.stdin)['count'])" 2>/dev/null || echo 0)
+if [ "$COUNT" -ge 1 ]; then
+  pass "Pattern filter session:* returned $COUNT event(s)"
+else
+  fail "Pattern filter session:* should return >= 1"
+fi
+
+# Since filter
+PAST=$(date -u -v-1H +"%Y-%m-%dT%H:%M:%SZ" 2>/dev/null || date -u -d "1 hour ago" +"%Y-%m-%dT%H:%M:%SZ" 2>/dev/null || echo "")
+if [ -n "$PAST" ]; then
+  RESP=$(curl -s "$BASE/api/events?since=$PAST&limit=10" -H "$AUTH")
+  COUNT=$(echo "$RESP" | python3 -c "import sys,json; print(json.load(sys.stdin)['count'])" 2>/dev/null || echo 0)
+  if [ "$COUNT" -ge 4 ]; then
+    pass "Since filter returned $COUNT recent events"
+  else
+    fail "Since filter should return >= 4 recent events"
+  fi
+fi
+
+# ─── 4. SSE Streaming ────────────────────────────────────────────────
+echo -e "\n${YELLOW}[4] SSE Real-Time Streaming${NC}"
+
+SSE_OUT=$(mktemp)
+curl -s -N -m 3 "$BASE/api/events/stream?patterns=test:*" -H "$AUTH" > "$SSE_OUT" 2>&1 &
+SSE_PID=$!
+sleep 0.5
+
+# Publish while SSE is listening
+curl -s -X POST "$BASE/api/events" \
+  -H "$AUTH" -H "Content-Type: application/json" \
+  -d '{"type":"custom","source":"sse-test","data":{"live":"stream"}}' > /dev/null
+
+# This won't match test:* pattern, publish one that does
+curl -s -X POST "$BASE/api/events" \
+  -H "$AUTH" -H "Content-Type: application/json" \
+  -d '{"type":"custom","source":"test:probe","data":{"sse":"verify"}}' > /dev/null
+
+sleep 2
+wait $SSE_PID 2>/dev/null || true
+
+if grep -q "retry:" "$SSE_OUT"; then
+  pass "SSE connection established (retry header)"
+else
+  fail "SSE did not send retry header"
+fi
+rm -f "$SSE_OUT"
+
+# ─── 5. WebSocket ────────────────────────────────────────────────────
+echo -e "\n${YELLOW}[5] WebSocket Streaming${NC}"
+
+if command -v wscat &>/dev/null; then
+  WS_OUT=$(mktemp)
+  wscat -c "ws://localhost:$PORT/ws/events?patterns=*&token=$TOKEN" --wait 2 > "$WS_OUT" 2>&1 &
+  WS_PID=$!
+  sleep 1
+
+  curl -s -X POST "$BASE/api/events" \
+    -H "$AUTH" -H "Content-Type: application/json" \
+    -d '{"type":"custom","source":"ws-test","data":{"ws":"verify"}}' > /dev/null
+
+  sleep 1
+  kill $WS_PID 2>/dev/null || true
+  wait $WS_PID 2>/dev/null || true
+
+  if grep -q "ws-test" "$WS_OUT"; then
+    pass "WebSocket received event"
+  else
+    fail "WebSocket did not receive event"
+  fi
+  rm -f "$WS_OUT"
+else
+  skip "WebSocket test (install: npm i -g wscat)"
+fi
+
+# ─── 6. Peter Parker Config ──────────────────────────────────────────
+echo -e "\n${YELLOW}[6] Peter Parker Config Generation${NC}"
+
+for PROFILE in cost balanced quality; do
+  FLAG="--$PROFILE"
+  node dist/index.js peter setup $FLAG 2>/dev/null
+  if [ -f .ralph-tui/config.toml ] && grep -q "Generated by Peter Parker" .ralph-tui/config.toml; then
+    AGENT_COUNT=$(grep -c '^\[\[agents\]\]' .ralph-tui/config.toml)
+    if [ "$AGENT_COUNT" -eq 10 ]; then
+      pass "$PROFILE profile: 10 agents generated"
+    else
+      fail "$PROFILE profile: expected 10 agents, got $AGENT_COUNT"
+    fi
+  else
+    fail "$PROFILE profile: config not generated"
+  fi
+done
+
+# Verify builder is default
+if grep -A3 'name = "builder"' .ralph-tui/config.toml | grep -q 'default = true'; then
+  pass "Builder agent is marked as default"
+else
+  fail "Builder should be default agent"
+fi
+
+# ─── 7. Peter Parker Status ──────────────────────────────────────────
+echo -e "\n${YELLOW}[7] Peter Parker Status${NC}"
+
+OUTPUT=$(node dist/index.js peter status 2>&1)
+if echo "$OUTPUT" | grep -q "Config profile"; then
+  pass "Peter status shows config profile"
+else
+  fail "Peter status missing config profile"
+fi
+
+if echo "$OUTPUT" | grep -q "Recent events"; then
+  pass "Peter status shows recent events"
+else
+  fail "Peter status missing recent events"
+fi
+
+# ─── 8. Event Persistence ────────────────────────────────────────────
+echo -e "\n${YELLOW}[8] Event Persistence${NC}"
+
+if [ -f .jfl/map-events.jsonl ]; then
+  LINE_COUNT=$(wc -l < .jfl/map-events.jsonl | tr -d ' ')
+  if [ "$LINE_COUNT" -ge 4 ]; then
+    pass "map-events.jsonl has $LINE_COUNT persisted events"
+  else
+    fail "Expected >= 4 persisted events, got $LINE_COUNT"
+  fi
+else
+  fail "map-events.jsonl not created"
+fi
+
+# Restart and check events survive
+node dist/index.js context-hub stop 2>/dev/null
+sleep 1
+node dist/index.js context-hub start 2>/dev/null
+sleep 2
+
+# Token regenerates on restart — re-read it
+TOKEN=$(cat .jfl/context-hub.token 2>/dev/null)
+AUTH="Authorization: Bearer $TOKEN"
+
+RESP=$(curl -s "$BASE/api/events?limit=50" -H "$AUTH" 2>/dev/null)
+COUNT=$(echo "$RESP" | python3 -c "import sys,json; print(json.load(sys.stdin)['count'])" 2>/dev/null || echo 0)
+if [ "$COUNT" -ge 4 ]; then
+  pass "Events survived restart ($COUNT recovered)"
+else
+  fail "Events lost on restart (got $COUNT)"
+fi
+
+# ─── 9. Pi CLI Detection ─────────────────────────────────────────────
+echo -e "\n${YELLOW}[9] Pi CLI Detection${NC}"
+
+if grep -q '"pi"' src/commands/session.ts && grep -q '"--yolo"' src/commands/session.ts; then
+  pass "Pi provider in session.ts"
+else
+  fail "Pi provider missing from session.ts"
+fi
+
+if command -v pi &>/dev/null; then
+  pass "Pi CLI found in PATH"
+else
+  skip "Pi CLI not installed (detection code is correct)"
+fi
+
+# ─── 10. Ralph Integration ───────────────────────────────────────────
+echo -e "\n${YELLOW}[10] Ralph Integration${NC}"
+
+if grep -q "PeterParkerBridge" src/commands/ralph.ts && grep -q "\-\-listen" src/commands/ralph.ts; then
+  pass "Ralph injects --listen and starts bridge"
+else
+  fail "Ralph missing --listen injection or bridge"
+fi
+
+if command -v ralph-tui &>/dev/null; then
+  pass "ralph-tui found in PATH"
+else
+  skip "ralph-tui not installed (integration code is correct)"
+fi
+
+# ─── Summary ─────────────────────────────────────────────────────────
+echo -e "\n${CYAN}═══════════════════════════════════════════════════════${NC}"
+echo -e "  ${GREEN}Passed: $PASS${NC}  ${RED}Failed: $FAIL${NC}  ${YELLOW}Skipped: $SKIP${NC}"
+echo -e "${CYAN}═══════════════════════════════════════════════════════${NC}\n"
+
+if [ "$FAIL" -gt 0 ]; then
+  exit 1
+fi

package/scripts/test-onboarding.sh

@@ -0,0 +1,121 @@
+#!/bin/bash
+# Test JFL onboarding flow in isolated environment
+
+set -e
+
+TEST_DIR="/tmp/jfl-test-$(date +%s)"
+CONFIG_BACKUP="$HOME/.config/jfl.backup-$(date +%s)"
+
+echo "🧪 JFL Test Mode"
+echo ""
+echo "This will:"
+echo "  1. Backup your current JFL config"
+echo "  2. Clear config for fresh onboarding"
+echo "  3. Create test directory: $TEST_DIR"
+echo "  4. Launch JFL"
+echo "  5. Restore your config when done"
+echo ""
+read -p "Continue? (y/n) " -n 1 -r
+echo
+if [[ ! $REPLY =~ ^[Yy]$ ]]; then
+    echo "Cancelled."
+    exit 0
+fi
+
+# Backup current config
+if [ -d "$HOME/.config/jfl" ]; then
+    echo "📦 Backing up config to $CONFIG_BACKUP"
+    mv "$HOME/.config/jfl" "$CONFIG_BACKUP"
+fi
+
+# Create test directory
+echo "📁 Creating test directory: $TEST_DIR"
+mkdir -p "$TEST_DIR"
+cd "$TEST_DIR"
+
+# Cleanup function
+cleanup() {
+    echo ""
+    echo "🧹 Cleaning up..."
+
+    # Check for created GitHub repos in test directory
+    CREATED_REPOS=()
+    if [ -d "$TEST_DIR" ]; then
+        # Find any directories with .git in them (product repos)
+        for dir in "$TEST_DIR"/*/product; do
+            if [ -d "$dir/.git" ]; then
+                # Get repo name from git remote
+                cd "$dir" 2>/dev/null || continue
+                REPO_URL=$(git remote get-url origin 2>/dev/null || echo "")
+                if [[ -n "$REPO_URL" ]]; then
+                    # Extract owner/repo from URL
+                    # Handles both SSH (git@github.com:owner/repo.git) and HTTPS (https://github.com/owner/repo.git)
+                    REPO_FULL=$(echo "$REPO_URL" | sed -E 's#^(https://|git@)github\.com[:/]##' | sed 's/\.git$//')
+                    CREATED_REPOS+=("$REPO_FULL")
+                fi
+            fi
+        done
+    fi
+
+    # Ask about deleting GitHub repos
+    if [ ${#CREATED_REPOS[@]} -gt 0 ]; then
+        echo ""
+        echo "Found ${#CREATED_REPOS[@]} test repo(s) on GitHub:"
+        for repo in "${CREATED_REPOS[@]}"; do
+            echo "  - $repo"
+        done
+        echo ""
+        read -p "Delete these repos from GitHub? (y/n) " -n 1 -r
+        echo
+        if [[ $REPLY =~ ^[Yy]$ ]]; then
+            for repo in "${CREATED_REPOS[@]}"; do
+                echo "Deleting $repo from GitHub..."
+                if gh repo delete "$repo" --yes 2>/dev/null; then
+                    echo "  ✓ Deleted: $repo"
+                else
+                    echo "  ⚠️  Failed to delete: $repo (may need to delete manually)"
+                fi
+            done
+        else
+            echo "Repos kept on GitHub"
+        fi
+    fi
+
+    # Remove test config
+    rm -rf "$HOME/.config/jfl"
+
+    # Restore original config
+    if [ -d "$CONFIG_BACKUP" ]; then
+        echo "📦 Restoring config from $CONFIG_BACKUP"
+        mv "$CONFIG_BACKUP" "$HOME/.config/jfl"
+    fi
+
+    # Ask about test directory
+    echo ""
+    read -p "Delete test directory $TEST_DIR? (y/n) " -n 1 -r
+    echo
+    if [[ $REPLY =~ ^[Yy]$ ]]; then
+        rm -rf "$TEST_DIR"
+        echo "✓ Test directory deleted"
+    else
+        echo "Test directory kept at: $TEST_DIR"
+    fi
+
+    echo ""
+    echo "✓ Restored to normal state"
+}
+
+# Set trap to cleanup on exit
+trap cleanup EXIT INT TERM
+
+# Show environment
+echo ""
+echo "🎯 Test Environment Ready"
+echo "   Config: Fresh (backed up to $CONFIG_BACKUP)"
+echo "   Directory: $TEST_DIR"
+echo ""
+echo "Starting JFL..."
+echo ""
+
+# Launch JFL
+jfl

package/scripts/test-planning-loop-e2e.ts

@@ -0,0 +1,181 @@
+/**
+ * End-to-end test for the Planning Loop
+ *
+ * Exercises: PolicyHead v2 → DynamicsModel rollouts → InvariantMonitor → Action Selection
+ *
+ * Run: npx tsx scripts/test-planning-loop-e2e.ts
+ */
+
+import { createPlanningLoop } from "../src/lib/planning-loop.js"
+import { PolicyHeadInference } from "../src/lib/policy-head.js"
+import type { PlanningResult, EvaluatedAction } from "../src/lib/planning-loop.js"
+
+const projectRoot = process.cwd()
+
+// ============================================================================
+// Test scenarios — represent real situations Peter Parker faces
+// ============================================================================
+
+const SCENARIOS = [
+  {
+    name: "Bug reported — tests failing",
+    agentId: "error-fixer",
+    goal: "Tests are failing in planning-loop.test.ts — TypeError on undefined property. Fix the failing test.",
+    expectedTypes: ["fix"],
+  },
+  {
+    name: "Feature request — add new capability",
+    agentId: "feature-builder",
+    goal: "Add multi-step rollout support to the planning loop for deeper lookahead.",
+    expectedTypes: ["feature"],
+  },
+  {
+    name: "Performance optimization",
+    agentId: "optimizer",
+    goal: "Reduce PolicyHead inference latency from 2s to under 500ms for interactive use.",
+    expectedTypes: ["experiment", "refactor"],
+  },
+  {
+    name: "Test coverage gap",
+    agentId: "test-coverage",
+    goal: "Add unit tests for the counterfactual training bridge — currently 0% coverage.",
+    expectedTypes: ["test"],
+  },
+  {
+    name: "Config change needed",
+    agentId: "config-updater",
+    goal: "Update the nightly pipeline schedule to run at 2am MST instead of midnight.",
+    expectedTypes: ["config"],
+  },
+]
+
+// ============================================================================
+// Helpers
+// ============================================================================
+
+function formatAction(ea: EvaluatedAction): string {
+  const pred = ea.prediction
+  const delta = pred.outcome?.immediate?.evalScoreChange
+  const deltaStr = delta !== undefined ? `Δ=${delta > 0 ? "+" : ""}${delta.toFixed(4)}` : "Δ=N/A"
+  const violations = ea.invariantViolations.length > 0 ? ` ⚠️${ea.invariantViolations.length} violations` : ""
+  const filtered = ea.filtered ? ` [FILTERED: ${ea.filterReason}]` : ""
+
+  return `  ${ea.action.actionType.padEnd(12)} conf=${ea.phConfidence.toFixed(3)} score=${ea.combinedScore.toFixed(3)} ${deltaStr} src=${pred.source}${violations}${filtered}`
+}
+
+function logResult(scenario: typeof SCENARIOS[0], result: PlanningResult) {
+  console.log(`\n${"═".repeat(70)}`)
+  console.log(`  Scenario: ${scenario.name}`)
+  console.log(`  Agent:    ${scenario.agentId}`)
+  console.log(`  Goal:     ${scenario.goal.slice(0, 70)}...`)
+  console.log(`${"─".repeat(70)}`)
+  console.log(`  Time:     ${result.planningTimeMs}ms`)
+  console.log(`  Rollouts: ${result.rolloutsPerformed}`)
+  console.log(`  Complete: ${result.completed}`)
+  console.log()
+
+  // All actions
+  console.log(`  Actions evaluated (${result.allActions.length}):`)
+  for (const ea of result.allActions) {
+    const selected = result.selectedAction === ea ? " ← SELECTED" : ""
+    console.log(`${formatAction(ea)}${selected}`)
+  }
+
+  // Selected action
+  if (result.selectedAction) {
+    const sel = result.selectedAction
+    const correctType = scenario.expectedTypes.includes(sel.action.actionType)
+    const check = correctType ? "✅" : "❌"
+    console.log(`\n  ${check} Selected: ${sel.action.actionType} (expected: ${scenario.expectedTypes.join("|")})`)
+  } else {
+    console.log(`\n  ❌ No action selected: ${result.noSelectionReason}`)
+  }
+}
+
+// ============================================================================
+// Main
+// ============================================================================
+
+async function main() {
+  console.log("╔══════════════════════════════════════════════════════════════════════╗")
+  console.log("║  Planning Loop — End-to-End Test                                    ║")
+  console.log("║  PH (v2 transformer) → DM (rollouts) → IM (invariants) → Select    ║")
+  console.log("╚══════════════════════════════════════════════════════════════════════╝")
+
+  // Check prerequisites
+  const ph = new PolicyHeadInference(projectRoot)
+  console.log(`\n  PolicyHead loaded: ${ph.isLoaded} (version: ${ph.version})`)
+  if (ph.stats) {
+    console.log(`  Trained on: ${ph.stats.trained_on} examples`)
+    const valAcc = ph.stats.val_accuracy ?? ph.stats.direction_accuracy
+    console.log(`  Val accuracy: ${(valAcc > 1 ? valAcc : valAcc * 100).toFixed(1)}%`)
+  }
+
+  if (!ph.isLoaded) {
+    console.error("\n  ❌ PolicyHead not loaded — cannot run e2e test")
+    console.error("  Copy checkpoint to .jfl/checkpoints/best_policy_head.pt")
+    process.exit(1)
+  }
+
+  // Create planning loop
+  const planner = createPlanningLoop(projectRoot, {
+    topK: 5,
+    verbose: true,
+    maxPlanningTimeMs: 30000, // 30s for test
+    checkInvariants: true,
+    recordTransitions: false, // Don't pollute real data during test
+    writeTrainingTuples: false,
+  })
+
+  console.log(`  Planning loop ready: ${planner.isReady}`)
+
+  // Run each scenario
+  let passed = 0
+  let failed = 0
+  const results: Array<{ scenario: string; result: PlanningResult; correct: boolean }> = []
+
+  for (const scenario of SCENARIOS) {
+    try {
+      console.log(`\n  Running: ${scenario.name}...`)
+      const result = await planner.plan(scenario.agentId, scenario.goal)
+      logResult(scenario, result)
+
+      const correct = result.selectedAction
+        ? scenario.expectedTypes.includes(result.selectedAction.action.actionType)
+        : false
+
+      if (correct) passed++
+      else failed++
+
+      results.push({ scenario: scenario.name, result, correct })
+    } catch (err: any) {
+      console.error(`\n  ❌ ${scenario.name} THREW: ${err.message}`)
+      failed++
+      results.push({
+        scenario: scenario.name,
+        result: null as any,
+        correct: false,
+      })
+    }
+  }
+
+  // Summary
+  console.log(`\n${"═".repeat(70)}`)
+  console.log(`  SUMMARY`)
+  console.log(`${"─".repeat(70)}`)
+  for (const r of results) {
+    const check = r.correct ? "✅" : "❌"
+    const time = r.result ? `${r.result.planningTimeMs}ms` : "ERRORED"
+    const selected = r.result?.selectedAction?.action.actionType ?? "none"
+    console.log(`  ${check} ${r.scenario.padEnd(35)} → ${selected.padEnd(12)} (${time})`)
+  }
+  console.log(`\n  Passed: ${passed}/${SCENARIOS.length}  Failed: ${failed}/${SCENARIOS.length}`)
+  console.log(`${"═".repeat(70)}`)
+
+  process.exit(failed > 0 ? 1 : 0)
+}
+
+main().catch(err => {
+  console.error("Fatal error:", err)
+  process.exit(1)
+})