claude-flow-novice 2.9.0 → 2.10.0

package/.claude/skills/cfn-redis-coordination/demos/test-iteration-feedback.sh

@@ -0,0 +1,320 @@
+#!/bin/bash
+#
+# Test Script: Iteration Feedback Mechanism
+#
+# Validates that:
+# 1. Validators can report feedback with confidence scores
+# 2. Orchestrator collects and aggregates feedback
+# 3. Feedback is passed to Loop 3 agents via wake signal
+# 4. Feedback is stored in Redis with TTL
+# 5. Agents can read feedback for iteration N
+
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+TASK_ID="test-feedback-$(date +%s)"
+
+echo "=========================================="
+echo "Iteration Feedback Mechanism Test"
+echo "=========================================="
+echo "Task ID: $TASK_ID"
+echo ""
+
+# Cleanup function
+cleanup() {
+  echo ""
+  echo "Cleaning up Redis keys..."
+  redis-cli --scan --pattern "swarm:${TASK_ID}:*" | xargs -r redis-cli del >/dev/null 2>&1 || true
+  echo "✅ Cleanup complete"
+}
+
+# Register cleanup on exit
+trap cleanup EXIT
+
+# Test 1: Validators report confidence with feedback
+echo "=== Test 1: Validators Report Feedback ==="
+echo ""
+
+VALIDATOR_1="reviewer-1"
+VALIDATOR_2="tester-1"
+VALIDATOR_3="security-1"
+
+# Validator 1: Low confidence with specific feedback
+echo "[$VALIDATOR_1] Reporting confidence 0.65 with 2 feedback items..."
+$SCRIPT_DIR/invoke-waiting-mode.sh report \
+  --task-id "$TASK_ID" \
+  --agent-id "$VALIDATOR_1" \
+  --confidence 0.65 \
+  --iteration 1 \
+  --feedback "Add error handling for null inputs,Improve documentation for API endpoints"
+
+echo ""
+
+# Validator 2: Medium confidence with 3 feedback items
+echo "[$VALIDATOR_2] Reporting confidence 0.72 with 3 feedback items..."
+$SCRIPT_DIR/invoke-waiting-mode.sh report \
+  --task-id "$TASK_ID" \
+  --agent-id "$VALIDATOR_2" \
+  --confidence 0.72 \
+  --iteration 1 \
+  --feedback "Increase test coverage to 80%,Add edge case tests for empty arrays,Fix flaky timeout test"
+
+echo ""
+
+# Validator 3: Low confidence with security feedback
+echo "[$VALIDATOR_3] Reporting confidence 0.60 with 2 feedback items..."
+$SCRIPT_DIR/invoke-waiting-mode.sh report \
+  --task-id "$TASK_ID" \
+  --agent-id "$VALIDATOR_3" \
+  --confidence 0.60 \
+  --iteration 1 \
+  --feedback "Fix SQL injection vulnerability in query builder,Add rate limiting to API endpoints"
+
+echo ""
+
+# Test 2: Collect feedback from validators
+echo "=== Test 2: Orchestrator Collects Feedback ==="
+echo ""
+
+COLLECT_OUTPUT=$($SCRIPT_DIR/invoke-waiting-mode.sh collect \
+  --task-id "$TASK_ID" \
+  --agent-ids "$VALIDATOR_1,$VALIDATOR_2,$VALIDATOR_3")
+
+echo "$COLLECT_OUTPUT"
+echo ""
+
+# Extract consensus and feedback
+CONSENSUS=$(echo "$COLLECT_OUTPUT" | tail -1)
+echo "Calculated Consensus: $CONSENSUS"
+
+# Verify feedback aggregation
+if echo "$COLLECT_OUTPUT" | grep -q "Aggregated Feedback"; then
+  echo "✅ Feedback aggregated successfully"
+  FEEDBACK_COUNT=$(echo "$COLLECT_OUTPUT" | grep -c '^\s*-' || echo "0")
+  echo "   Total feedback items: $FEEDBACK_COUNT"
+else
+  echo "❌ No aggregated feedback found"
+  exit 1
+fi
+
+echo ""
+
+# Test 3: Wake Loop 3 agent with validator feedback
+echo "=== Test 3: Wake Loop 3 Agent with Feedback ==="
+echo ""
+
+LOOP3_AGENT="backend-dev"
+
+# Extract aggregated feedback for passing to Loop 3
+LOOP2_FEEDBACK=$(echo "$COLLECT_OUTPUT" | sed -n '/Aggregated Feedback/,/Consensus:/p' | grep '^\s*-' | sed 's/^\s*-\s*//' | paste -sd ',' -)
+
+# Build full feedback message
+FULL_FEEDBACK="Improve consensus from $CONSENSUS to >=0.90,$LOOP2_FEEDBACK"
+
+echo "Waking $LOOP3_AGENT with comprehensive feedback..."
+echo "Feedback items: $(echo "$FULL_FEEDBACK" | tr ',' '\n' | wc -l)"
+echo ""
+
+$SCRIPT_DIR/invoke-waiting-mode.sh wake \
+  --task-id "$TASK_ID" \
+  --agent-id "$LOOP3_AGENT" \
+  --reason "cfn_loop_iteration" \
+  --iteration 2 \
+  --priority 30 \
+  --feedback "$FULL_FEEDBACK"
+
+echo ""
+
+# Test 4: Verify feedback stored in Redis
+echo "=== Test 4: Verify Feedback Storage ==="
+echo ""
+
+FEEDBACK_KEY="swarm:${TASK_ID}:${LOOP3_AGENT}:feedback:iteration-2"
+echo "Checking Redis key: $FEEDBACK_KEY"
+
+STORED_FEEDBACK=$(redis-cli get "$FEEDBACK_KEY")
+
+if [ -n "$STORED_FEEDBACK" ] && [ "$STORED_FEEDBACK" != "(nil)" ]; then
+  echo "✅ Feedback stored successfully"
+  echo ""
+  echo "Stored feedback (JSON array):"
+  echo "$STORED_FEEDBACK" | jq '.'
+  echo ""
+
+  # Verify TTL
+  TTL=$(redis-cli ttl "$FEEDBACK_KEY")
+  echo "TTL: $TTL seconds ($(echo "scale=1; $TTL / 3600" | bc) hours)"
+
+  if [ "$TTL" -gt 0 ] && [ "$TTL" -le 86400 ]; then
+    echo "✅ TTL is correct (≤24 hours)"
+  else
+    echo "❌ TTL is incorrect: $TTL"
+    exit 1
+  fi
+else
+  echo "❌ Feedback not found in Redis"
+  exit 1
+fi
+
+echo ""
+
+# Test 5: Agent reads feedback
+echo "=== Test 5: Agent Reads Feedback ==="
+echo ""
+
+echo "Simulating agent reading feedback from Redis..."
+AGENT_FEEDBACK=$(redis-cli get "$FEEDBACK_KEY")
+
+if [ -n "$AGENT_FEEDBACK" ] && [ "$AGENT_FEEDBACK" != "(nil)" ]; then
+  echo "✅ Agent successfully read feedback"
+  echo ""
+  echo "Feedback items available to agent:"
+  echo "$AGENT_FEEDBACK" | jq -r '.[]' | nl -w2 -s'. '
+  echo ""
+
+  ITEM_COUNT=$(echo "$AGENT_FEEDBACK" | jq '. | length')
+  echo "Total actionable items: $ITEM_COUNT"
+else
+  echo "❌ Agent failed to read feedback"
+  exit 1
+fi
+
+echo ""
+
+# Test 6: Wake queue contains feedback
+echo "=== Test 6: Verify Wake Queue Feedback ==="
+echo ""
+
+WAKE_QUEUE="swarm:${TASK_ID}:${LOOP3_AGENT}:wake-queue"
+echo "Checking wake queue: $WAKE_QUEUE"
+
+# Pop wake message from queue
+WAKE_MSG=$(redis-cli ZPOPMIN "$WAKE_QUEUE" 1 | sed -n '1p')
+
+if [ -n "$WAKE_MSG" ] && [ "$WAKE_MSG" != "(nil)" ]; then
+  echo "✅ Wake message found in queue"
+  echo ""
+  echo "Wake message contents:"
+  echo "$WAKE_MSG" | jq '.'
+  echo ""
+
+  # Verify feedback in wake message
+  WAKE_FEEDBACK=$(echo "$WAKE_MSG" | jq -r '.feedback[]' 2>/dev/null | head -3)
+  if [ -n "$WAKE_FEEDBACK" ]; then
+    echo "✅ Feedback included in wake message"
+    echo ""
+    echo "First 3 feedback items:"
+    echo "$WAKE_MSG" | jq -r '.feedback[]' | head -3 | nl -w2 -s'. '
+  else
+    echo "❌ No feedback in wake message"
+    exit 1
+  fi
+else
+  echo "❌ No wake message in queue"
+  exit 1
+fi
+
+echo ""
+
+# Test 7: Multiple iterations
+echo "=== Test 7: Multiple Iteration Feedback ==="
+echo ""
+
+echo "Simulating iteration 3 feedback..."
+ITERATION_3_FEEDBACK="Address remaining issues from iteration 2,Add integration tests,Update changelog"
+
+$SCRIPT_DIR/invoke-waiting-mode.sh wake \
+  --task-id "$TASK_ID" \
+  --agent-id "$LOOP3_AGENT" \
+  --reason "cfn_loop_iteration" \
+  --iteration 3 \
+  --priority 30 \
+  --feedback "$ITERATION_3_FEEDBACK"
+
+echo ""
+
+# Verify both iteration feedbacks are stored separately
+FEEDBACK_KEY_2="swarm:${TASK_ID}:${LOOP3_AGENT}:feedback:iteration-2"
+FEEDBACK_KEY_3="swarm:${TASK_ID}:${LOOP3_AGENT}:feedback:iteration-3"
+
+FEEDBACK_2=$(redis-cli get "$FEEDBACK_KEY_2")
+FEEDBACK_3=$(redis-cli get "$FEEDBACK_KEY_3")
+
+if [ -n "$FEEDBACK_2" ] && [ "$FEEDBACK_2" != "(nil)" ]; then
+  echo "✅ Iteration 2 feedback persists"
+  echo "   Items: $(echo "$FEEDBACK_2" | jq '. | length')"
+fi
+
+if [ -n "$FEEDBACK_3" ] && [ "$FEEDBACK_3" != "(nil)" ]; then
+  echo "✅ Iteration 3 feedback stored"
+  echo "   Items: $(echo "$FEEDBACK_3" | jq '. | length')"
+fi
+
+echo ""
+
+# Test 8: Edge cases
+echo "=== Test 8: Edge Cases ==="
+echo ""
+
+# Empty feedback
+echo "Testing wake with empty feedback..."
+$SCRIPT_DIR/invoke-waiting-mode.sh wake \
+  --task-id "$TASK_ID" \
+  --agent-id "test-agent" \
+  --reason "test" \
+  --iteration 1 \
+  --priority 50 \
+  --feedback ""
+
+EMPTY_FEEDBACK_KEY="swarm:${TASK_ID}:test-agent:feedback:iteration-1"
+EMPTY_RESULT=$(redis-cli get "$EMPTY_FEEDBACK_KEY")
+
+if [ -z "$EMPTY_RESULT" ] || [ "$EMPTY_RESULT" = "(nil)" ]; then
+  echo "✅ Empty feedback not stored (expected)"
+else
+  echo "⚠️  Empty feedback stored: $EMPTY_RESULT"
+fi
+
+echo ""
+
+# Iteration 0 (no feedback storage)
+echo "Testing wake with iteration 0..."
+$SCRIPT_DIR/invoke-waiting-mode.sh wake \
+  --task-id "$TASK_ID" \
+  --agent-id "test-agent-2" \
+  --reason "initial" \
+  --iteration 0 \
+  --priority 50 \
+  --feedback "This should not be stored"
+
+ITER_0_KEY="swarm:${TASK_ID}:test-agent-2:feedback:iteration-0"
+ITER_0_RESULT=$(redis-cli get "$ITER_0_KEY")
+
+if [ -z "$ITER_0_RESULT" ] || [ "$ITER_0_RESULT" = "(nil)" ]; then
+  echo "✅ Iteration 0 feedback not stored (expected)"
+else
+  echo "❌ Iteration 0 feedback should not be stored: '$ITER_0_RESULT'"
+  exit 1
+fi
+
+echo ""
+
+# Summary
+echo "=========================================="
+echo "Test Summary"
+echo "=========================================="
+echo "✅ All tests passed!"
+echo ""
+echo "Validated:"
+echo "  1. Validators can report feedback with confidence"
+echo "  2. Orchestrator collects and aggregates feedback"
+echo "  3. Feedback passed to Loop 3 agents via wake signal"
+echo "  4. Feedback stored in Redis with 24-hour TTL"
+echo "  5. Agents can read feedback for iteration N"
+echo "  6. Wake queue contains feedback in message"
+echo "  7. Multiple iterations maintain separate feedback"
+echo "  8. Edge cases handled correctly"
+echo ""
+echo "Task ID: $TASK_ID"
+echo "Feedback mechanism is fully operational!"
+echo ""

package/.claude/skills/cfn-redis-coordination/demos/test-orchestrator.sh

@@ -0,0 +1,249 @@
+#!/bin/bash
+
+# Test Suite for orchestrate-cfn-loop.sh
+# Tests dependency enforcement, BLPOP blocking, and agent completion protocol
+
+set -e
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+ORCHESTRATOR="$SCRIPT_DIR/orchestrate-cfn-loop.sh"
+
+# Colors for output
+GREEN='\033[0;32m'
+RED='\033[0;31m'
+YELLOW='\033[1;33m'
+NC='\033[0m' # No Color
+
+# Test counters
+TESTS_RUN=0
+TESTS_PASSED=0
+TESTS_FAILED=0
+
+# Test helpers
+function test_start() {
+  echo -e "${YELLOW}[TEST]${NC} $1"
+  TESTS_RUN=$((TESTS_RUN + 1))
+}
+
+function test_pass() {
+  echo -e "${GREEN}  ✓ PASS${NC}"
+  TESTS_PASSED=$((TESTS_PASSED + 1))
+}
+
+function test_fail() {
+  echo -e "${RED}  ✗ FAIL${NC} $1"
+  TESTS_FAILED=$((TESTS_FAILED + 1))
+}
+
+function cleanup_redis() {
+  local TASK_ID="$1"
+  redis-cli --scan --pattern "swarm:${TASK_ID}:*" | xargs -r redis-cli del > /dev/null 2>&1 || true
+}
+
+# ==============================================================================
+# TEST 1: Orchestrator script exists and is executable
+# ==============================================================================
+test_start "Orchestrator script exists and is executable"
+
+if [ -f "$ORCHESTRATOR" ] && [ -x "$ORCHESTRATOR" ]; then
+  test_pass
+else
+  test_fail "Script not found or not executable: $ORCHESTRATOR"
+fi
+
+# ==============================================================================
+# TEST 2: Redis connection available
+# ==============================================================================
+test_start "Redis connection available"
+
+if redis-cli ping > /dev/null 2>&1; then
+  test_pass
+else
+  test_fail "Redis not available. Start Redis server: redis-server"
+  exit 1
+fi
+
+# ==============================================================================
+# TEST 3: Orchestrator script is a bash script
+# ==============================================================================
+test_start "Orchestrator script is a bash script"
+
+# Check if orchestrator exists (validation will be added later)
+if [ -f "$ORCHESTRATOR" ]; then
+  test_pass
+else
+  test_fail "Orchestrator script not found"
+fi
+
+# ==============================================================================
+# TEST 4: Agent completion protocol - Signal :done key
+# ==============================================================================
+test_start "Agent completion protocol - Signal :done key"
+
+TASK_ID="test-completion-$(date +%s)"
+AGENT_ID="test-agent-1"
+
+cleanup_redis "$TASK_ID"
+
+# Simulate agent completing work
+redis-cli lpush "swarm:${TASK_ID}:${AGENT_ID}:done" "complete" > /dev/null
+
+# Verify signal present
+SIGNAL=$(redis-cli blpop "swarm:${TASK_ID}:${AGENT_ID}:done" 1 2>&1 | tail -n 1)
+if [ "$SIGNAL" == "complete" ]; then
+  test_pass
+else
+  test_fail "Signal not found in Redis"
+fi
+
+cleanup_redis "$TASK_ID"
+
+# ==============================================================================
+# TEST 5: BLPOP blocking behavior (timeout)
+# ==============================================================================
+test_start "BLPOP blocking behavior (timeout)"
+
+TASK_ID="test-blocking-$(date +%s)"
+AGENT_ID="nonexistent-agent"
+
+cleanup_redis "$TASK_ID"
+
+# BLPOP with 1 second timeout (should timeout since no agent signals)
+START_TIME=$(date +%s)
+redis-cli blpop "swarm:${TASK_ID}:${AGENT_ID}:done" 1 > /dev/null 2>&1
+END_TIME=$(date +%s)
+ELAPSED=$((END_TIME - START_TIME))
+
+if [ "$ELAPSED" -ge 1 ] && [ "$ELAPSED" -le 2 ]; then
+  test_pass
+else
+  test_fail "BLPOP timeout behavior incorrect (elapsed: ${ELAPSED}s, expected: 1-2s)"
+fi
+
+cleanup_redis "$TASK_ID"
+
+# ==============================================================================
+# TEST 6: Agent completion protocol - Report confidence
+# ==============================================================================
+test_start "Agent completion protocol - Report confidence"
+
+TASK_ID="test-confidence-$(date +%s)"
+AGENT_ID="test-agent-1"
+
+cleanup_redis "$TASK_ID"
+
+# Simulate agent reporting confidence
+OUTPUT=$(./.claude/skills/cfn-redis-coordination/invoke-waiting-mode.sh report \
+  --task-id "$TASK_ID" \
+  --agent-id "$AGENT_ID" \
+  --confidence 0.85 \
+  --iteration 1 2>&1)
+
+# Check if report was successful
+if echo "$OUTPUT" | grep -q "0.85" || redis-cli exists "swarm:${TASK_ID}:${AGENT_ID}:result" | grep -q "1"; then
+  test_pass
+else
+  test_fail "Confidence not stored correctly"
+fi
+
+cleanup_redis "$TASK_ID"
+
+# ==============================================================================
+# TEST 7: Collect consensus from multiple agents
+# ==============================================================================
+test_start "Collect consensus from multiple agents"
+
+TASK_ID="test-consensus-$(date +%s)"
+
+cleanup_redis "$TASK_ID"
+
+# Simulate 3 agents reporting confidence
+./.claude/skills/cfn-redis-coordination/invoke-waiting-mode.sh report \
+  --task-id "$TASK_ID" \
+  --agent-id "agent-1" \
+  --confidence 0.85 \
+  --iteration 1 > /dev/null 2>&1
+
+./.claude/skills/cfn-redis-coordination/invoke-waiting-mode.sh report \
+  --task-id "$TASK_ID" \
+  --agent-id "agent-2" \
+  --confidence 0.90 \
+  --iteration 1 > /dev/null 2>&1
+
+./.claude/skills/cfn-redis-coordination/invoke-waiting-mode.sh report \
+  --task-id "$TASK_ID" \
+  --agent-id "agent-3" \
+  --confidence 0.78 \
+  --iteration 1 > /dev/null 2>&1
+
+# Collect consensus
+CONSENSUS=$(./.claude/skills/cfn-redis-coordination/invoke-waiting-mode.sh collect \
+  --task-id "$TASK_ID" \
+  --agent-ids "agent-1,agent-2,agent-3" 2>&1)
+
+# Check for .84 or 0.84 in output (average: (0.85 + 0.90 + 0.78) / 3 = 0.843)
+if echo "$CONSENSUS" | grep -qE "(0\.84|\.84)"; then
+  test_pass
+else
+  test_fail "Consensus calculation incorrect: $CONSENSUS"
+fi
+
+cleanup_redis "$TASK_ID"
+
+# ==============================================================================
+# TEST 8: Mode-specific thresholds documentation
+# ==============================================================================
+test_start "Mode-specific thresholds documentation"
+
+# Check if CLAUDE.md has mode thresholds (primary documentation location)
+if grep -q "MVP" /mnt/c/Users/masha/Documents/claude-flow-novice/CLAUDE.md 2>/dev/null && \
+   grep -q "Standard" /mnt/c/Users/masha/Documents/claude-flow-novice/CLAUDE.md 2>/dev/null; then
+  test_pass
+else
+  test_fail "Mode-specific thresholds should be documented (reference: CLAUDE.md)"
+fi
+
+# ==============================================================================
+# TEST 9: Deliverable verification (BUG #11 fix)
+# ==============================================================================
+test_start "Deliverable verification - prevents consensus on vapor"
+
+TASK_ID="deliverable-test-$(date +%s)"
+
+# Simulate Loop 3 completion with high confidence but NO files changed
+redis-cli lpush "swarm:${TASK_ID}:coder-1:done" "complete" > /dev/null 2>&1
+redis-cli lpush "swarm:${TASK_ID}:coder-1:result" "0.95" > /dev/null 2>&1
+
+# Verify orchestrator checks git status and overrides confidence when no files exist
+# The implementation should check: git status --short | grep -E "^(A|M|\?\?)" | wc -l
+# If count == 0, override confidence to 0.0
+
+# Check that deliverable verification logic exists in orchestrator
+if grep -q "FILES_CHANGED.*git status.*short" /mnt/c/Users/masha/Documents/claude-flow-novice/.claude/skills/cfn-redis-coordination/orchestrate-cfn-loop.sh && \
+   grep -q "no_deliverables\|DELIVERABLE.*FAILED" /mnt/c/Users/masha/Documents/claude-flow-novice/.claude/skills/cfn-redis-coordination/orchestrate-cfn-loop.sh; then
+  test_pass
+else
+  test_fail "Deliverable verification logic not found in orchestrator"
+fi
+
+cleanup_redis "$TASK_ID"
+
+# ==============================================================================
+# TEST SUMMARY
+# ==============================================================================
+echo ""
+echo "========================================"
+echo "TEST SUMMARY"
+echo "========================================"
+echo -e "Tests Run:    ${TESTS_RUN}"
+echo -e "Tests Passed: ${GREEN}${TESTS_PASSED}${NC}"
+echo -e "Tests Failed: ${RED}${TESTS_FAILED}${NC}"
+echo "========================================"
+
+if [ "$TESTS_FAILED" -eq 0 ]; then
+  echo -e "${GREEN}✓ All tests passed${NC}"
+  exit 0
+else
+  echo -e "${RED}✗ Some tests failed${NC}"
+  exit 1
+fi