qualia-framework 6.14.0 → 6.22.0

package/tests/eval-runner.test.sh

@@ -0,0 +1,147 @@
+#!/bin/bash
+# eval-runner.test.sh — bin/eval-runner.js (layered AI-feature eval, R7)
+# Run: bash tests/eval-runner.test.sh
+
+PASS=0
+FAIL=0
+BIN_DIR="$(cd "$(dirname "$0")/../bin" && pwd)"
+NODE="${NODE:-node}"
+ER="$BIN_DIR/eval-runner.js"
+
+assert_exit() {
+  local name="$1" expected="$2" actual="$3"
+  if [ "$expected" = "$actual" ]; then echo "  ✓ $name"; PASS=$((PASS+1));
+  else echo "  ✗ $name (expected exit $expected, got $actual)"; FAIL=$((FAIL+1)); fi
+}
+assert_contains() {
+  local name="$1" hay="$2" needle="$3"
+  if echo "$hay" | grep -qF "$needle"; then echo "  ✓ $name"; PASS=$((PASS+1));
+  else echo "  ✗ $name (missing '$needle' in: $hay)"; FAIL=$((FAIL+1)); fi
+}
+
+echo "eval-runner.test.sh — bin/eval-runner.js"
+echo ""
+
+$NODE -c "$ER" 2>/dev/null && { echo "  ✓ syntax valid"; PASS=$((PASS+1)); } || { echo "  ✗ syntax invalid"; FAIL=$((FAIL+1)); }
+
+# --- all deterministic assertions pass → PASS ---
+TMP=$(mktemp -d)
+cat > "$TMP/s.json" <<'EOF'
+{ "feature":"support-chat", "cases":[
+  { "name":"refund", "output":"We refund within 30 days.", "latency_ms":1200, "cost_usd":0.008,
+    "assert":[
+      {"type":"contains","value":"30 days"},
+      {"type":"not_contains","value":"I cannot help"},
+      {"type":"regex","value":"\\b30\\b"},
+      {"type":"min_length","value":5},
+      {"type":"max_latency_ms","value":2000},
+      {"type":"max_cost_usd","value":0.02}
+    ] } ] }
+EOF
+$NODE "$ER" "$TMP/s.json" >/dev/null 2>&1
+assert_exit "all deterministic pass → exit 0" 0 $?
+OUT=$($NODE "$ER" "$TMP/s.json" 2>&1)
+assert_contains "reports EVAL PASS" "$OUT" "EVAL PASS"
+rm -rf "$TMP"
+
+# --- a failing contains → FAIL with the failing assertion shown ---
+TMP=$(mktemp -d)
+cat > "$TMP/s.json" <<'EOF'
+{ "feature":"chat", "cases":[
+  { "name":"refusal", "output":"I cannot help with that.",
+    "assert":[ {"type":"contains","value":"30 days"}, {"type":"not_contains","value":"I cannot help"} ] } ] }
+EOF
+$NODE "$ER" "$TMP/s.json" >/dev/null 2>&1
+assert_exit "failing assertion → exit 1" 1 $?
+OUT=$($NODE "$ER" "$TMP/s.json" 2>&1)
+assert_contains "shows failing case" "$OUT" "refusal"
+rm -rf "$TMP"
+
+# --- latency over budget → FAIL ---
+TMP=$(mktemp -d)
+cat > "$TMP/s.json" <<'EOF'
+{ "feature":"chat", "cases":[
+  { "name":"slow", "output":"ok", "latency_ms":5000, "assert":[ {"type":"max_latency_ms","value":2000} ] } ] }
+EOF
+$NODE "$ER" "$TMP/s.json" >/dev/null 2>&1
+assert_exit "latency over budget → exit 1" 1 $?
+# missing latency metric but asserted → fail (no silent pass)
+cat > "$TMP/s2.json" <<'EOF'
+{ "feature":"chat", "cases":[ { "name":"no-metric", "output":"ok", "assert":[ {"type":"max_latency_ms","value":2000} ] } ] }
+EOF
+$NODE "$ER" "$TMP/s2.json" >/dev/null 2>&1
+assert_exit "asserting latency with none recorded → exit 1" 1 $?
+rm -rf "$TMP"
+
+# --- json_path + json_valid ---
+TMP=$(mktemp -d)
+cat > "$TMP/s.json" <<'EOF'
+{ "feature":"rag", "cases":[
+  { "name":"structured", "output":"{\"answer\":\"yes\",\"sources\":[{\"id\":\"doc1\"}]}",
+    "assert":[
+      {"type":"json_valid"},
+      {"type":"json_path","path":"answer","equals":"yes"},
+      {"type":"json_path","path":"sources.0.id","contains":"doc"}
+    ] } ] }
+EOF
+$NODE "$ER" "$TMP/s.json" >/dev/null 2>&1
+assert_exit "json_path + json_valid pass → exit 0" 0 $?
+# invalid json under json_valid → fail
+cat > "$TMP/bad.json" <<'EOF'
+{ "feature":"rag", "cases":[ { "name":"notjson", "output":"not json", "assert":[ {"type":"json_valid"} ] } ] }
+EOF
+$NODE "$ER" "$TMP/bad.json" >/dev/null 2>&1
+assert_exit "json_valid on non-json → exit 1" 1 $?
+rm -rf "$TMP"
+
+# --- llm_rubric: verdict pass / fail / pending ---
+TMP=$(mktemp -d)
+cat > "$TMP/pass.json" <<'EOF'
+{ "feature":"chat", "cases":[ { "name":"r", "output":"grounded answer", "assert":[ {"type":"llm_rubric","rubric":"grounded","verdict":"pass"} ] } ] }
+EOF
+$NODE "$ER" "$TMP/pass.json" >/dev/null 2>&1
+assert_exit "llm_rubric verdict=pass → exit 0" 0 $?
+cat > "$TMP/fail.json" <<'EOF'
+{ "feature":"chat", "cases":[ { "name":"r", "output":"hallucination", "assert":[ {"type":"llm_rubric","rubric":"grounded","verdict":"fail"} ] } ] }
+EOF
+$NODE "$ER" "$TMP/fail.json" >/dev/null 2>&1
+assert_exit "llm_rubric verdict=fail → exit 1" 1 $?
+cat > "$TMP/pending.json" <<'EOF'
+{ "feature":"chat", "cases":[ { "name":"r", "output":"x", "assert":[ {"type":"llm_rubric","rubric":"grounded"} ] } ] }
+EOF
+$NODE "$ER" "$TMP/pending.json" >/dev/null 2>&1
+assert_exit "unjudged llm_rubric → PENDING → exit 1 (no silent pass)" 1 $?
+OUT=$($NODE "$ER" "$TMP/pending.json" 2>&1)
+assert_contains "reports unjudged rubric count" "$OUT" "unjudged"
+rm -rf "$TMP"
+
+# --- output_file resolution + --write artifact ---
+TMP=$(mktemp -d); mkdir -p "$TMP/.planning"
+echo -n "the answer is 42" > "$TMP/out.txt"
+cat > "$TMP/s.json" <<'EOF'
+{ "feature":"file-feature", "cases":[ { "name":"fromfile", "output_file":"out.txt", "assert":[ {"type":"contains","value":"42"} ] } ] }
+EOF
+(cd "$TMP" && $NODE "$ER" s.json --write >/dev/null 2>&1)
+assert_exit "output_file case passes → exit 0" 0 $?
+[ -f "$TMP/.planning/evals/eval-file-feature.json" ] && { echo "  ✓ --write emits eval artifact"; PASS=$((PASS+1)); } || { echo "  ✗ no eval artifact"; FAIL=$((FAIL+1)); }
+# missing output_file → fail, not crash
+cat > "$TMP/miss.json" <<'EOF'
+{ "feature":"x", "cases":[ { "name":"gone", "output_file":"nope.txt", "assert":[ {"type":"contains","value":"y"} ] } ] }
+EOF
+(cd "$TMP" && $NODE "$ER" miss.json >/dev/null 2>&1)
+assert_exit "missing output_file → exit 1 (graceful)" 1 $?
+rm -rf "$TMP"
+
+# --- library: runAssertion unit ---
+U=$($NODE -e "const e=require('$ER'); console.log(e.runAssertion({type:'contains',value:'hi'},{output:'oh hi there'}).ok)" 2>&1)
+assert_contains "runAssertion contains true" "$U" "true"
+U=$($NODE -e "const e=require('$ER'); console.log(e.getPath({a:{b:[{c:9}]}},'a.b.0.c'))" 2>&1)
+assert_contains "getPath nested+index" "$U" "9"
+
+# --- malformed suite → exit 2 ---
+$NODE "$ER" /nonexistent-suite.json >/dev/null 2>&1
+assert_exit "missing suite → exit 2" 2 $?
+
+echo ""
+echo "=== Results: $PASS passed, $FAIL failed ==="
+[ "$FAIL" -eq 0 ] && exit 0 || exit 1

package/tests/instructions.test.sh

@@ -0,0 +1,109 @@
+#!/bin/bash
+# instructions.test.sh — single-source instruction compile (R4) + host adapter
+# contract (R5). Run: bash tests/instructions.test.sh
+
+PASS=0
+FAIL=0
+FRAMEWORK_DIR="$(cd "$(dirname "$0")/.." && pwd)"
+NODE="${NODE:-node}"
+HA="$FRAMEWORK_DIR/bin/host-adapters.js"
+CI="$FRAMEWORK_DIR/bin/compile-instructions.js"
+
+assert_exit() {
+  local name="$1" expected="$2" actual="$3"
+  if [ "$expected" = "$actual" ]; then echo "  ✓ $name"; PASS=$((PASS+1));
+  else echo "  ✗ $name (expected exit $expected, got $actual)"; FAIL=$((FAIL+1)); fi
+}
+assert_eq() {
+  local name="$1" expected="$2" actual="$3"
+  if [ "$expected" = "$actual" ]; then echo "  ✓ $name"; PASS=$((PASS+1));
+  else echo "  ✗ $name (expected '$expected', got '$actual')"; FAIL=$((FAIL+1)); fi
+}
+assert_contains() {
+  local name="$1" hay="$2" needle="$3"
+  if echo "$hay" | grep -qF "$needle"; then echo "  ✓ $name"; PASS=$((PASS+1));
+  else echo "  ✗ $name (missing '$needle')"; FAIL=$((FAIL+1)); fi
+}
+refute_contains() {
+  local name="$1" hay="$2" needle="$3"
+  if echo "$hay" | grep -qF "$needle"; then echo "  ✗ $name (found '$needle')"; FAIL=$((FAIL+1));
+  else echo "  ✓ $name"; PASS=$((PASS+1)); fi
+}
+
+echo "instructions.test.sh — single-source compile (R4) + host contract (R5)"
+echo ""
+
+# syntax
+$NODE -c "$HA" 2>/dev/null && { echo "  ✓ host-adapters syntax valid"; PASS=$((PASS+1)); } || { echo "  ✗ host-adapters syntax"; FAIL=$((FAIL+1)); }
+$NODE -c "$CI" 2>/dev/null && { echo "  ✓ compile-instructions syntax valid"; PASS=$((PASS+1)); } || { echo "  ✗ compile-instructions syntax"; FAIL=$((FAIL+1)); }
+
+# --- R4: drift guard — committed files match the canonical ---
+$NODE "$CI" --check >/dev/null 2>&1
+assert_exit "compile-instructions --check passes (no drift)" 0 $?
+
+# A canonical edit not reflected in the artifacts is caught by --check.
+TMPCANON=$(mktemp)
+cp "$FRAMEWORK_DIR/templates/instructions.md" "$TMPCANON"
+printf '\nDRIFT-MARKER\n' >> "$FRAMEWORK_DIR/templates/instructions.md"
+$NODE "$CI" --check >/dev/null 2>&1
+RC=$?
+cp "$TMPCANON" "$FRAMEWORK_DIR/templates/instructions.md"   # restore immediately
+rm -f "$TMPCANON"
+assert_exit "drift guard fails (exit 1) when canonical edited but not recompiled" 1 $RC
+# confirm restore worked
+$NODE "$CI" --check >/dev/null 2>&1
+assert_exit "canonical restored, --check green again" 0 $?
+
+# --- R4: both artifacts share an identical body, differ only in footer ---
+CLAUDE_BODY=$(sed -n '/# Qualia Framework/,/state router tells you the next command/p' "$FRAMEWORK_DIR/CLAUDE.md")
+AGENTS_BODY=$(sed -n '/# Qualia Framework/,/state router tells you the next command/p' "$FRAMEWORK_DIR/AGENTS.md")
+if [ "$CLAUDE_BODY" = "$AGENTS_BODY" ]; then
+  echo "  ✓ CLAUDE.md and AGENTS.md bodies are identical (no drift)"; PASS=$((PASS+1));
+else
+  echo "  ✗ CLAUDE.md and AGENTS.md bodies diverge"; FAIL=$((FAIL+1));
+fi
+# host-specific footer survives
+assert_contains "CLAUDE.md keeps Claude footer" "$(cat "$FRAMEWORK_DIR/CLAUDE.md")" "this file stays under 25 lines"
+assert_contains "AGENTS.md keeps cross-vendor footer" "$(cat "$FRAMEWORK_DIR/AGENTS.md")" "cross-vendor compatibility"
+refute_contains "AGENTS.md does NOT carry the Claude-only footer" "$(cat "$FRAMEWORK_DIR/AGENTS.md")" "this file stays under 25 lines"
+# both carry the generated header + role placeholder for install to fill
+assert_contains "CLAUDE.md has generated header" "$(head -1 "$FRAMEWORK_DIR/CLAUDE.md")" "GENERATED from templates/instructions.md"
+assert_contains "AGENTS.md keeps {{ROLE}} for install" "$(cat "$FRAMEWORK_DIR/AGENTS.md")" "{{ROLE}}"
+
+# --- R5: adapter contract is the single source of per-host facts ---
+assert_eq "claude instructionFile" "CLAUDE.md" "$($NODE -e "console.log(require('$HA').adapter('claude').instructionFile)")"
+assert_eq "codex instructionFile" "AGENTS.md" "$($NODE -e "console.log(require('$HA').adapter('codex').instructionFile)")"
+assert_eq "claude configFile" "settings.json" "$($NODE -e "console.log(require('$HA').adapter('claude').configFile)")"
+assert_eq "codex configFile" "config.toml" "$($NODE -e "console.log(require('$HA').adapter('codex').configFile)")"
+assert_eq "codex agentExt" ".toml" "$($NODE -e "console.log(require('$HA').adapter('codex').agentExt)")"
+assert_eq "unknown host throws" "throws" "$($NODE -e "try{require('$HA').adapter('cursor');console.log('no')}catch(e){console.log('throws')}")"
+
+# --- R5: stripHostBlocks keeps only the matching host's block ---
+SRC='before
+<!--QUALIA-HOST claude-->
+CLAUDE_ONLY
+<!--/QUALIA-HOST-->
+<!--QUALIA-HOST codex-->
+CODEX_ONLY
+<!--/QUALIA-HOST-->
+after'
+CLAUDE_OUT=$($NODE -e "console.log(require('$HA').stripHostBlocks(process.argv[1],'claude'))" "$SRC")
+assert_contains "claude strip keeps CLAUDE_ONLY" "$CLAUDE_OUT" "CLAUDE_ONLY"
+refute_contains "claude strip drops CODEX_ONLY" "$CLAUDE_OUT" "CODEX_ONLY"
+CODEX_OUT=$($NODE -e "console.log(require('$HA').stripHostBlocks(process.argv[1],'codex'))" "$SRC")
+assert_contains "codex strip keeps CODEX_ONLY" "$CODEX_OUT" "CODEX_ONLY"
+refute_contains "codex strip drops CLAUDE_ONLY" "$CODEX_OUT" "CLAUDE_ONLY"
+
+# --- R5: compileInstructions applies naming but leaves tokens/role for install ---
+NAMED=$($NODE -e "console.log(require('$HA').compileInstructions('Use Claude Code now. Role: {{ROLE}} at \${QUALIA_BIN}','codex'))")
+assert_contains "codex compile swaps the product name" "$NAMED" "Use Codex now"
+assert_contains "codex compile leaves {{ROLE}} for install" "$NAMED" "{{ROLE}}"
+assert_contains "codex compile leaves \${QUALIA_BIN} token for install" "$NAMED" '${QUALIA_BIN}'
+
+# --- R5: renderText still resolves paths (regression) ---
+RENDERED=$($NODE -e "console.log(require('$HA').renderText('bin at \${QUALIA_BIN}','claude'))")
+assert_contains "renderText resolves QUALIA_BIN" "$RENDERED" "/.claude/bin"
+
+echo ""
+echo "=== Results: $PASS passed, $FAIL failed ==="
+[ "$FAIL" -eq 0 ] && exit 0 || exit 1

package/tests/last-report.test.sh

@@ -0,0 +1,156 @@
+#!/bin/bash
+# last-report.test.sh — bin/last-report.js (surface the latest session report)
+# Run: bash tests/last-report.test.sh
+
+PASS=0
+FAIL=0
+BIN_DIR="$(cd "$(dirname "$0")/../bin" && pwd)"
+NODE="${NODE:-node}"
+LR="$BIN_DIR/last-report.js"
+
+assert_exit() {
+  local name="$1" expected="$2" actual="$3"
+  if [ "$expected" = "$actual" ]; then echo "  ✓ $name"; PASS=$((PASS+1));
+  else echo "  ✗ $name (expected exit $expected, got $actual)"; FAIL=$((FAIL+1)); fi
+}
+assert_contains() {
+  local name="$1" hay="$2" needle="$3"
+  if echo "$hay" | grep -qF "$needle"; then echo "  ✓ $name"; PASS=$((PASS+1));
+  else echo "  ✗ $name (missing '$needle' in: $hay)"; FAIL=$((FAIL+1)); fi
+}
+assert_not_contains() {
+  local name="$1" hay="$2" needle="$3"
+  if echo "$hay" | grep -qF "$needle"; then echo "  ✗ $name (unexpected '$needle' in: $hay)"; FAIL=$((FAIL+1));
+  else echo "  ✓ $name"; PASS=$((PASS+1)); fi
+}
+
+# Write a well-formed report at $1/.planning/reports/report-$2.md with summary $3, next $4
+write_report() {
+  local root="$1" datepart="$2" summary="$3" next="$4" date="$5"
+  mkdir -p "$root/.planning/reports"
+  cat > "$root/.planning/reports/report-$datepart.md" <<EOF
+# Session Report — $date
+
+**Project:** demo
+**Employee:** T
+**Branch:** main
+**Date:** $date
+
+## What Was Done
+- $summary
+
+## Blockers
+None.
+
+## Next Steps
+1. $next
+
+## Commits (1)
+- abc123 seed
+EOF
+}
+
+echo "last-report.test.sh — bin/last-report.js"
+echo ""
+
+$NODE -c "$LR" 2>/dev/null && { echo "  ✓ syntax valid"; PASS=$((PASS+1)); } || { echo "  ✗ syntax invalid"; FAIL=$((FAIL+1)); }
+
+# --- no reports dir → exit 1, found:false ---
+TMP=$(mktemp -d)
+(cd "$TMP" && $NODE "$LR" >/dev/null 2>&1)
+assert_exit "no reports dir → exit 1" 1 $?
+OUT=$(cd "$TMP" && $NODE "$LR" --json 2>&1)
+assert_contains "json found:false when none" "$OUT" '"found": false'
+OUT=$(cd "$TMP" && $NODE "$LR" 2>&1)
+assert_contains "human: nothing to surface" "$OUT" "No session reports"
+rm -rf "$TMP"
+
+# --- single report → extracts date + summary + next, exit 0 ---
+TMP=$(mktemp -d)
+write_report "$TMP" "2026-05-28" "Built the login flow" "Wire the dashboard" "2026-05-28"
+(cd "$TMP" && $NODE "$LR" >/dev/null 2>&1)
+assert_exit "single report → exit 0" 0 $?
+OUT=$(cd "$TMP" && $NODE "$LR" 2>&1)
+assert_contains "human shows date" "$OUT" "2026-05-28"
+assert_contains "human shows summary" "$OUT" "Built the login flow"
+assert_contains "human shows next" "$OUT" "Wire the dashboard"
+assert_contains "human has 'Last session' prefix" "$OUT" "Last session"
+rm -rf "$TMP"
+
+# --- json shape ---
+TMP=$(mktemp -d)
+write_report "$TMP" "2026-05-28" "Built the login flow" "Wire the dashboard" "2026-05-28"
+OUT=$(cd "$TMP" && $NODE "$LR" --json 2>&1)
+assert_contains "json found:true" "$OUT" '"found": true'
+assert_contains "json file field" "$OUT" '"file": "report-2026-05-28.md"'
+assert_contains "json date field" "$OUT" '"date": "2026-05-28"'
+assert_contains "json summary field" "$OUT" '"summary": "Built the login flow"'
+assert_contains "json next field" "$OUT" '"next": "Wire the dashboard"'
+rm -rf "$TMP"
+
+# --- multiple reports → picks the newest BY DATE (not file order) ---
+TMP=$(mktemp -d)
+write_report "$TMP" "2026-05-28" "Older work" "Old next" "2026-05-28"
+write_report "$TMP" "2026-06-15" "Newest work" "New next" "2026-06-15"
+write_report "$TMP" "2026-06-01" "Middle work" "Mid next" "2026-06-01"
+OUT=$(cd "$TMP" && $NODE "$LR" --json 2>&1)
+assert_contains "picks newest date file" "$OUT" '"file": "report-2026-06-15.md"'
+assert_contains "picks newest summary" "$OUT" "Newest work"
+assert_not_contains "does not pick older summary" "$OUT" "Older work"
+rm -rf "$TMP"
+
+# --- dated suffix filenames (report-2026-06-20-session2.md) parse correctly ---
+TMP=$(mktemp -d)
+write_report "$TMP" "2026-06-10" "Earlier" "e" "2026-06-10"
+write_report "$TMP" "2026-06-20-session2" "Suffixed newest" "s" "2026-06-20"
+OUT=$(cd "$TMP" && $NODE "$LR" --json 2>&1)
+assert_contains "suffixed filename wins by date" "$OUT" "Suffixed newest"
+assert_contains "suffixed date extracted" "$OUT" '"date": "2026-06-20"'
+rm -rf "$TMP"
+
+# --- --now makes age_days deterministic ---
+TMP=$(mktemp -d)
+write_report "$TMP" "2026-06-01" "Work" "Next" "2026-06-01"
+OUT=$(cd "$TMP" && $NODE "$LR" --json --now 2026-06-11T00:00:00Z 2>&1)
+assert_contains "age_days deterministic (10)" "$OUT" '"age_days": 10'
+OUT=$(cd "$TMP" && $NODE "$LR" --now 2026-06-11T00:00:00Z 2>&1)
+assert_contains "human age uses --now" "$OUT" "10d ago"
+rm -rf "$TMP"
+
+# --- library: latestReport() returns structured object ---
+TMP=$(mktemp -d)
+write_report "$TMP" "2026-06-01" "Lib work" "Lib next" "2026-06-01"
+RES=$($NODE -e "console.log(require('$LR').latestReport('$TMP').summary)" 2>&1)
+assert_contains "latestReport() returns summary" "$RES" "Lib work"
+RES=$($NODE -e "console.log(require('$LR').latestReport('$TMP',{now:'2026-06-06T00:00:00Z'}).age_days)" 2>&1)
+assert_contains "latestReport() honors now opt (5)" "$RES" "5"
+rm -rf "$TMP"
+
+# --- malformed / empty report handled gracefully (no crash, still exit 0) ---
+TMP=$(mktemp -d)
+mkdir -p "$TMP/.planning/reports"
+: > "$TMP/.planning/reports/report-2026-06-02.md"   # completely empty
+(cd "$TMP" && $NODE "$LR" >/dev/null 2>&1)
+assert_exit "empty report → exit 0 (found by filename)" 0 $?
+OUT=$(cd "$TMP" && $NODE "$LR" --json 2>&1)
+assert_contains "empty report found:true" "$OUT" '"found": true'
+assert_contains "empty report date from filename" "$OUT" '"date": "2026-06-02"'
+rm -rf "$TMP"
+
+# --- malformed report (no headings, just prose) still yields a summary ---
+TMP=$(mktemp -d)
+mkdir -p "$TMP/.planning/reports"
+printf 'just some freeform notes about the day\n' > "$TMP/.planning/reports/report-2026-06-03.md"
+OUT=$(cd "$TMP" && $NODE "$LR" --json 2>&1)
+assert_contains "malformed report falls back to first line" "$OUT" "just some freeform notes"
+rm -rf "$TMP"
+
+# --- bad argument → exit 2 ---
+TMP=$(mktemp -d)
+(cd "$TMP" && $NODE "$LR" --bogus >/dev/null 2>&1)
+assert_exit "unknown flag → exit 2" 2 $?
+rm -rf "$TMP"
+
+echo ""
+echo "=== Results: $PASS passed, $FAIL failed ==="
+[ "$FAIL" -eq 0 ] && exit 0 || exit 1

package/tests/lib.test.sh

@@ -507,7 +507,7 @@ TMP=$(mktmp)
 mkdir -p "$TMP/home/.claude/bin" "$TMP/home/.claude/hooks" "$TMP/home/.claude/knowledge/daily-log" "$TMP/home/.claude/qualia-design" "$TMP/home/.claude/agents" "$TMP/home/.claude/qualia-templates" "$TMP/project"
 echo '{"installed_by":"Test","role":"OWNER","version":"6.3.0","erp":{"enabled":false}}' > "$TMP/home/.claude/.qualia-config.json"
 touch "$TMP/home/.claude/CLAUDE.md" "$TMP/home/.claude/settings.json"
-for f in runtime-manifest.js command-surface.js host-adapters.js state.js qualia-ui.js statusline.js knowledge.js knowledge-flush.js state-ledger.js plan-contract.js contract-runner.js agent-status.js analyze-gate.js harness-eval.js trust-score.js agent-runs.js slop-detect.mjs erp-retry.js work-packet.js report-payload.js project-snapshot.js codex-goal.js planning-hygiene.js prune-deprecated.js learning-candidates.js status-snapshot.js security-scan.js auto-report.js; do
+for f in runtime-manifest.js command-surface.js host-adapters.js state.js qualia-ui.js statusline.js knowledge.js knowledge-flush.js state-ledger.js plan-contract.js contract-runner.js agent-status.js analyze-gate.js verify-panel.js wave-plan.js eval-runner.js branch-hygiene.js last-report.js harness-eval.js trust-score.js agent-runs.js slop-detect.mjs erp-retry.js work-packet.js report-payload.js project-snapshot.js project-sync.js codex-goal.js planning-hygiene.js prune-deprecated.js learning-candidates.js status-snapshot.js security-scan.js auto-report.js; do
   touch "$TMP/home/.claude/bin/$f"
 done
 for h in session-start.js auto-update.js branch-guard.js pre-push.js pre-deploy-gate.js migration-guard.js git-guardrails.js stop-session-log.js fawzi-approval-guard.js vercel-account-guard.js env-empty-guard.js supabase-destructive-guard.js; do
@@ -622,7 +622,7 @@ TMP=$(mktmp)
 mkdir -p "$TMP/.claude/bin" "$TMP/.claude/hooks" "$TMP/.claude/knowledge/daily-log" "$TMP/.claude/qualia-design" "$TMP/.claude/agents" "$TMP/.claude/qualia-templates" "$TMP/project/.planning"
 echo '{"installed_by":"Test","role":"OWNER","erp":{"enabled":false}}' > "$TMP/.claude/.qualia-config.json"
 touch "$TMP/.claude/CLAUDE.md" "$TMP/.claude/settings.json"
-for f in runtime-manifest.js command-surface.js host-adapters.js state.js qualia-ui.js statusline.js knowledge.js knowledge-flush.js state-ledger.js plan-contract.js contract-runner.js agent-status.js analyze-gate.js harness-eval.js trust-score.js agent-runs.js slop-detect.mjs erp-retry.js work-packet.js report-payload.js project-snapshot.js codex-goal.js planning-hygiene.js prune-deprecated.js learning-candidates.js status-snapshot.js security-scan.js auto-report.js; do
+for f in runtime-manifest.js command-surface.js host-adapters.js state.js qualia-ui.js statusline.js knowledge.js knowledge-flush.js state-ledger.js plan-contract.js contract-runner.js agent-status.js analyze-gate.js verify-panel.js wave-plan.js eval-runner.js branch-hygiene.js last-report.js harness-eval.js trust-score.js agent-runs.js slop-detect.mjs erp-retry.js work-packet.js report-payload.js project-snapshot.js project-sync.js codex-goal.js planning-hygiene.js prune-deprecated.js learning-candidates.js status-snapshot.js security-scan.js auto-report.js; do
   touch "$TMP/.claude/bin/$f"
 done
 for h in session-start.js auto-update.js branch-guard.js pre-push.js pre-deploy-gate.js migration-guard.js git-guardrails.js stop-session-log.js fawzi-approval-guard.js vercel-account-guard.js env-empty-guard.js supabase-destructive-guard.js; do

package/tests/project-sync.test.sh

@@ -0,0 +1,175 @@
+#!/bin/bash
+# project-sync.test.sh — bin/project-sync.js (Framework→ERP project-sync payload)
+# Run: bash tests/project-sync.test.sh
+
+PASS=0
+FAIL=0
+BIN_DIR="$(cd "$(dirname "$0")/../bin" && pwd)"
+NODE="${NODE:-node}"
+PS="$BIN_DIR/project-sync.js"
+
+assert_exit() {
+  local name="$1" expected="$2" actual="$3"
+  if [ "$expected" = "$actual" ]; then echo "  ✓ $name"; PASS=$((PASS+1));
+  else echo "  ✗ $name (expected exit $expected, got $actual)"; FAIL=$((FAIL+1)); fi
+}
+assert_contains() {
+  local name="$1" hay="$2" needle="$3"
+  if echo "$hay" | grep -qF "$needle"; then echo "  ✓ $name"; PASS=$((PASS+1));
+  else echo "  ✗ $name (missing '$needle' in output)"; FAIL=$((FAIL+1)); fi
+}
+assert_not_contains() {
+  local name="$1" hay="$2" needle="$3"
+  if echo "$hay" | grep -qF "$needle"; then echo "  ✗ $name (unexpected '$needle')"; FAIL=$((FAIL+1));
+  else echo "  ✓ $name"; PASS=$((PASS+1)); fi
+}
+
+# A fully-populated .planning fixture: milestones, REQ-IDs, offroad, lifetime.
+setup_full() {
+  local tmp
+  tmp=$(mktemp -d)
+  mkdir -p "$tmp/.planning"
+  cat > "$tmp/.planning/tracking.json" <<'EOF'
+{
+  "project": "acme-portal",
+  "project_id": "qs-acme-portal",
+  "client": "Acme",
+  "milestone": 2,
+  "milestone_name": "Product",
+  "milestones": [
+    { "num": 1, "name": "Foundation", "closed_at": "2026-04-10T18:00:00Z", "phases_completed": 3, "tasks_completed": 12, "deployed_url": "https://m1.vercel.app" }
+  ],
+  "phase": 2,
+  "phase_name": "Dashboard",
+  "total_phases": 4,
+  "status": "built",
+  "tasks_done": 3,
+  "tasks_total": 5,
+  "verification": "pending",
+  "gap_cycles": { "2": 1 },
+  "build_count": 4,
+  "deploy_count": 1,
+  "deployed_url": "https://client.vercel.app",
+  "lifecycle": "build",
+  "lifetime": { "tasks_completed": 15, "phases_completed": 4, "milestones_completed": 1, "total_phases": 7, "offroad_count": 2 },
+  "offroad": [
+    { "at": "2026-06-01T10:00:00Z", "milestone": 2, "ref": "BUG-7", "note": "hotfix login" },
+    { "at": "2026-06-02T10:00:00Z", "milestone": 2, "ref": null, "note": "tweak" }
+  ]
+}
+EOF
+  cat > "$tmp/.planning/JOURNEY.md" <<'EOF'
+## Milestone 1 · Foundation
+## Milestone 2 · Product [CURRENT]
+## Milestone 3 · Handoff [FINAL]
+EOF
+  cat > "$tmp/.planning/REQUIREMENTS.md" <<'EOF'
+| ID | Milestone | Phase | Status |
+|----|-----------|-------|--------|
+| CORE-01 | M1: Foundation | Phase 1 | Complete |
+| CORE-02 | M2: Product | Phase 2 | Complete |
+| CORE-03 | M2: Product | Phase 3 | Incomplete |
+EOF
+  echo "$tmp"
+}
+
+echo "project-sync.test.sh — bin/project-sync.js"
+echo ""
+
+# --- syntax ---
+$NODE -c "$PS" 2>/dev/null && { echo "  ✓ syntax valid"; PASS=$((PASS+1)); } || { echo "  ✗ syntax invalid"; FAIL=$((FAIL+1)); }
+
+NOW="2026-06-21T00:00:00.000Z"
+
+# --- no .planning → exit 2 ---
+TMP=$(mktemp -d)
+(cd "$TMP" && $NODE "$PS" >/dev/null 2>&1)
+assert_exit "no .planning → exit 2" 2 $?
+rm -rf "$TMP"
+
+# --- full fixture → exit 0 + valid JSON ---
+TMP=$(setup_full)
+OUT=$($NODE "$PS" --cwd "$TMP" --json --now "$NOW" 2>&1)
+(cd "$TMP" && $NODE "$PS" --json --now "$NOW" >/dev/null 2>&1)
+assert_exit "full fixture → exit 0" 0 $?
+echo "$OUT" | $NODE -e "let s='';process.stdin.on('data',d=>s+=d).on('end',()=>{JSON.parse(s);process.exit(0)})" 2>/dev/null \
+  && { echo "  ✓ --json emits valid JSON"; PASS=$((PASS+1)); } \
+  || { echo "  ✗ --json invalid JSON"; FAIL=$((FAIL+1)); }
+
+# --- schema_version + payload marker ---
+assert_contains "carries schema_version" "$OUT" '"schema_version": 1'
+assert_contains "payload marker is project-sync" "$OUT" '"payload": "project-sync"'
+
+# --- lifecycle + launched_at handling ---
+assert_contains "lifecycle present (build)" "$OUT" '"lifecycle": "build"'
+
+# --- milestones[] with status + REQ completion ---
+assert_contains "milestone 1 closed" "$OUT" '"status": "closed"'
+assert_contains "milestone 2 current" "$OUT" '"status": "current"'
+assert_contains "milestone 3 future" "$OUT" '"status": "future"'
+assert_contains "closed milestone phases count" "$OUT" '"phases": 3'
+assert_contains "closed milestone tasks_completed" "$OUT" '"tasks_completed": 12'
+assert_contains "REQ total tracked for M2" "$OUT" '"total": 2'
+assert_contains "REQ complete count" "$OUT" '"complete": 1'
+assert_contains "incomplete REQ id surfaced" "$OUT" '"id": "CORE-03"'
+assert_contains "future milestone REQ untracked" "$OUT" '"tracked": false'
+assert_contains "per-milestone deployed_url" "$OUT" '"deployed_url": "https://m1.vercel.app"'
+assert_contains "total_milestones from journey" "$OUT" '"total_milestones": 3'
+
+# --- current position ---
+assert_contains "current phase" "$OUT" '"phase": 2'
+assert_contains "current verification" "$OUT" '"verification": "pending"'
+
+# --- task rollup ---
+assert_contains "rollup tasks_completed" "$OUT" '"tasks_completed": 15'
+assert_contains "rollup build_count" "$OUT" '"build_count": 4'
+assert_contains "rollup deploy_count" "$OUT" '"deploy_count": 1'
+assert_contains "rollup gap_cycles (flattened to number)" "$OUT" '"current_phase_gap_cycles": 1'
+
+# --- accountability / offroad ---
+assert_contains "offroad_count" "$OUT" '"offroad_count": 2'
+assert_contains "recent offroad entry" "$OUT" '"ref": "BUG-7"'
+
+# --- integration / merge model ---
+assert_contains "trunk integration model" "$OUT" '"model": "trunk"'
+assert_contains "integrates at ship" "$OUT" '"integrates_at": "/qualia-ship"'
+assert_contains "main-push event type" "$OUT" '"main_push_event_type": "employee_main_push"'
+
+rm -rf "$TMP"
+
+# --- graceful when fields absent: minimal tracking.json, no JOURNEY/REQUIREMENTS ---
+TMP=$(mktemp -d)
+mkdir -p "$TMP/.planning"
+echo '{ "project": "bare", "milestone": 1, "phase": 1 }' > "$TMP/.planning/tracking.json"
+OUT=$($NODE "$PS" --cwd "$TMP" --json --now "$NOW" 2>&1)
+(cd "$TMP" && $NODE "$PS" --json --now "$NOW" >/dev/null 2>&1)
+assert_exit "bare fixture → exit 0" 0 $?
+assert_contains "bare still has schema_version" "$OUT" '"schema_version": 1'
+assert_contains "bare offroad_count defaults to 0" "$OUT" '"offroad_count": 0'
+assert_contains "bare emits a current milestone" "$OUT" '"status": "current"'
+assert_contains "bare REQ untracked (no REQUIREMENTS.md)" "$OUT" '"tracked": false'
+rm -rf "$TMP"
+
+# --- library: buildProjectSync() returns the object ---
+TMP=$(setup_full)
+RES=$($NODE -e "const {buildProjectSync}=require('$PS');const o=buildProjectSync({cwd:'$TMP',now:'$NOW'});console.log(o.schema_version, o.milestones.length, o.milestones[1].requirements.incomplete[0].id)" 2>&1)
+assert_contains "buildProjectSync() returns enriched object" "$RES" "1 3 CORE-03"
+rm -rf "$TMP"
+
+# --- library: milestoneRequirements() parses a fixture dir ---
+TMP=$(setup_full)
+RES=$($NODE -e "const {milestoneRequirements}=require('$PS');const r=milestoneRequirements('$TMP/.planning',2);console.log(r.tracked, r.total, r.complete, r.incomplete.length)" 2>&1)
+assert_contains "milestoneRequirements() counts REQ completion" "$RES" "true 2 1 1"
+rm -rf "$TMP"
+
+# --- --write persists a snapshot file ---
+TMP=$(setup_full)
+FILE=$($NODE "$PS" --cwd "$TMP" --write --now "$NOW" 2>&1)
+if [ -f "$FILE" ]; then echo "  ✓ --write persists a file"; PASS=$((PASS+1));
+else echo "  ✗ --write did not persist ($FILE)"; FAIL=$((FAIL+1)); fi
+assert_contains "written file is project-sync-*" "$FILE" "project-sync-"
+rm -rf "$TMP"
+
+echo ""
+echo "=== Results: $PASS passed, $FAIL failed ==="
+[ "$FAIL" -eq 0 ] && exit 0 || exit 1

package/tests/run-all.sh

@@ -20,6 +20,13 @@ SUITES=(
   "slop-detect"
   "agent-status"
   "analyze-gate"
+  "instructions"
+  "verify-panel"
+  "wave-plan"
+  "eval-runner"
+  "branch-hygiene"
+  "last-report"
+  "project-sync"
 )
 
 FAILED=()