claude-flow-novice 2.10.5 → 2.10.7

package/claude-assets/skills/cfn-ace-system/invoke-context-reflect.sh

@@ -0,0 +1,343 @@
+#!/usr/bin/env bash
+
+##############################################################################
+# ACE Context Reflection Wrapper (Phase 3.1 - Anti-Pattern Detection)
+# Generates cognitive reflections on context data with failure analysis
+#
+# Usage:
+#   ./invoke-context-reflect.sh --context '{"task": "..."}' [OPTIONS]
+#
+# Arguments:
+#   --context          JSON context object (required for legacy mode)
+#   --confidence       Sprint confidence score 0.0-1.0 (for anti-pattern mode)
+#   --iterations       Number of iterations required (for anti-pattern mode)
+#   --feedback         ITERATE feedback from Product Owner (for anti-pattern mode)
+#   --task-id          Task/sprint identifier (required for anti-pattern mode)
+#   --swarm-id         Swarm identifier (optional)
+#   --sprint-ref       Sprint reference for tracking (optional)
+#   --domain           Domain classification (optional)
+#   --final-decision   Final Product Owner decision (PROCEED/ITERATE/ABORT)
+#   --final-feedback   Final iteration feedback (solution extraction)
+#   --complexity       Override complexity calculation (optional)
+#   --output           Output file path (optional, default: stdout)
+#   --memory-path      SQLite memory path (optional)
+##############################################################################
+
+set -euo pipefail
+
+# Default values
+CONTEXT=""
+CONFIDENCE=""
+ITERATIONS=""
+FEEDBACK=""
+TASK_ID=""
+SWARM_ID="default"
+SPRINT_REF=""
+DOMAIN=""
+FINAL_DECISION=""
+FINAL_FEEDBACK=""
+COMPLEXITY=""
+OUTPUT=""
+MEMORY_PATH="${ACE_MEMORY_PATH:-./.artifacts/database/swarm-memory.db}"
+
+# Parse arguments
+while [[ $# -gt 0 ]]; do
+  case $1 in
+    --context)
+      CONTEXT="$2"
+      shift 2
+      ;;
+    --confidence)
+      CONFIDENCE="$2"
+      shift 2
+      ;;
+    --iterations)
+      ITERATIONS="$2"
+      shift 2
+      ;;
+    --feedback)
+      FEEDBACK="$2"
+      shift 2
+      ;;
+    --task-id)
+      TASK_ID="$2"
+      shift 2
+      ;;
+    --swarm-id)
+      SWARM_ID="$2"
+      shift 2
+      ;;
+    --sprint-ref)
+      SPRINT_REF="$2"
+      shift 2
+      ;;
+    --domain)
+      DOMAIN="$2"
+      shift 2
+      ;;
+    --final-decision)
+      FINAL_DECISION="$2"
+      shift 2
+      ;;
+    --final-feedback)
+      FINAL_FEEDBACK="$2"
+      shift 2
+      ;;
+    --complexity)
+      COMPLEXITY="$2"
+      shift 2
+      ;;
+    --output)
+      OUTPUT="$2"
+      shift 2
+      ;;
+    --memory-path)
+      MEMORY_PATH="$2"
+      shift 2
+      ;;
+    *)
+      echo "Unknown option: $1"
+      exit 1
+      ;;
+  esac
+done
+
+# Get project root
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+PROJECT_ROOT="$(cd "$SCRIPT_DIR/../../.." && pwd)"
+
+# Extract failure reason from feedback
+extract_failure_reason() {
+  local feedback="$1"
+
+  # Parse common failure patterns
+  if echo "$feedback" | grep -qi "missing error"; then
+    echo "Missing error handling"
+  elif echo "$feedback" | grep -qi "security"; then
+    echo "Security vulnerability detected"
+  elif echo "$feedback" | grep -qi "test.*fail"; then
+    echo "Test failures"
+  elif echo "$feedback" | grep -qi "performance"; then
+    echo "Performance issues"
+  elif echo "$feedback" | grep -qi "missing.*validation"; then
+    echo "Input validation missing"
+  elif echo "$feedback" | grep -qi "no.*deliverable"; then
+    echo "No deliverables created"
+  elif echo "$feedback" | grep -qi "coverage"; then
+    echo "Insufficient test coverage"
+  elif echo "$feedback" | grep -qi "documentation"; then
+    echo "Missing documentation"
+  else
+    # Extract first sentence or first 100 chars
+    echo "$feedback" | head -n 1 | cut -c 1-100
+  fi
+}
+
+# Extract solution from final feedback
+extract_solution() {
+  local feedback="$1"
+
+  if echo "$feedback" | grep -qi "added.*error"; then
+    echo "Added comprehensive error handling"
+  elif echo "$feedback" | grep -qi "implemented.*security"; then
+    echo "Implemented security best practices"
+  elif echo "$feedback" | grep -qi "fixed.*test"; then
+    echo "Fixed test suite and increased coverage"
+  elif echo "$feedback" | grep -qi "optimized.*performance"; then
+    echo "Optimized performance bottlenecks"
+  else
+    # Return generic solution message
+    echo "Iterative improvements applied"
+  fi
+}
+
+# Generate tags from feedback
+generate_tags() {
+  local feedback="$1"
+  local tags="[]"
+
+  # Build tags array
+  if echo "$feedback" | grep -qi "error"; then
+    tags=$(echo "$tags" | jq '. + ["error-handling"]')
+  fi
+  if echo "$feedback" | grep -qi "security"; then
+    tags=$(echo "$tags" | jq '. + ["security"]')
+  fi
+  if echo "$feedback" | grep -qi "test"; then
+    tags=$(echo "$tags" | jq '. + ["testing"]')
+  fi
+  if echo "$feedback" | grep -qi "performance"; then
+    tags=$(echo "$tags" | jq '. + ["performance"]')
+  fi
+  if echo "$feedback" | grep -qi "validation"; then
+    tags=$(echo "$tags" | jq '. + ["validation"]')
+  fi
+  if echo "$feedback" | grep -qi "coverage"; then
+    tags=$(echo "$tags" | jq '. + ["coverage"]')
+  fi
+  if echo "$feedback" | grep -qi "documentation"; then
+    tags=$(echo "$tags" | jq '. + ["documentation"]')
+  fi
+  if echo "$feedback" | grep -qi "deliverable"; then
+    tags=$(echo "$tags" | jq '. + ["deliverable-tracking"]')
+  fi
+
+  echo "$tags"
+}
+
+# Anti-Pattern Detection Mode
+if [ -n "$CONFIDENCE" ] && [ -n "$ITERATIONS" ] && [ -n "$FEEDBACK" ] && [ -n "$TASK_ID" ]; then
+
+  # Determine lesson type and severity
+  LESSON_TYPE="strategy"
+  SEVERITY="info"
+
+  # Critical anti-pattern (confidence < 0.50)
+  if (( $(echo "$CONFIDENCE < 0.50" | bc -l) )); then
+    LESSON_TYPE="anti-pattern"
+    SEVERITY="critical"
+  # Warning (confidence < 0.70)
+  elif (( $(echo "$CONFIDENCE < 0.70" | bc -l) )); then
+    LESSON_TYPE="warning"
+    SEVERITY="warning"
+  # Low confidence success (0.70-0.75)
+  elif (( $(echo "$CONFIDENCE < 0.75" | bc -l) )); then
+    LESSON_TYPE="pattern"
+    SEVERITY="info"
+  fi
+
+  # Extract failure reason and solution
+  FAILURE_REASON=$(extract_failure_reason "$FEEDBACK")
+  SOLUTION=""
+  if [ -n "$FINAL_DECISION" ] && [ "$FINAL_DECISION" == "PROCEED" ] && [ -n "$FINAL_FEEDBACK" ]; then
+    SOLUTION=$(extract_solution "$FINAL_FEEDBACK")
+  fi
+
+  # Generate tags
+  TAGS=$(generate_tags "$FEEDBACK")
+  if [ -n "$DOMAIN" ]; then
+    TAGS=$(echo "$TAGS" | jq ". + [\"$DOMAIN\"]")
+  fi
+
+  # Build extracted lessons
+  LESSON_CONTENT=""
+  if [ "$LESSON_TYPE" == "anti-pattern" ]; then
+    LESSON_CONTENT="Avoid: $FAILURE_REASON (caused $ITERATIONS iterations, confidence $CONFIDENCE)"
+  elif [ "$LESSON_TYPE" == "warning" ]; then
+    LESSON_CONTENT="Caution: $FAILURE_REASON (required $ITERATIONS iterations)"
+  else
+    LESSON_CONTENT="Pattern observed: $FAILURE_REASON (confidence $CONFIDENCE after $ITERATIONS iterations)"
+  fi
+
+  # Add solution if available
+  EXTRACTED_LESSONS="{\"anti_pattern\": \"$LESSON_CONTENT\""
+  if [ -n "$SOLUTION" ]; then
+    EXTRACTED_LESSONS="$EXTRACTED_LESSONS, \"solution\": \"$SOLUTION\""
+  fi
+  EXTRACTED_LESSONS="$EXTRACTED_LESSONS}"
+
+  # Build metadata
+  METADATA="{\"tags\": $TAGS, \"severity\": \"$SEVERITY\""
+  if [ -n "$DOMAIN" ]; then
+    METADATA="$METADATA, \"domain\": \"$DOMAIN\""
+  fi
+  if [ -n "$SPRINT_REF" ]; then
+    METADATA="$METADATA, \"sprint_ref\": \"$SPRINT_REF\""
+  fi
+  METADATA="$METADATA, \"failure_reason\": \"$FAILURE_REASON\"}"
+
+  # Generate unique ID
+  REFLECTION_ID="ref-$(date +%s)-$(echo $RANDOM | md5sum | head -c 8)"
+
+  # Store in SQLite
+  sqlite3 "$MEMORY_PATH" <<EOF
+INSERT INTO context_reflections (
+  id, reflection_type, task_id, swarm_id,
+  execution_trace, feedback_signals, extracted_lessons,
+  metadata, confidence, created_at
+) VALUES (
+  '$REFLECTION_ID',
+  '$LESSON_TYPE',
+  '$TASK_ID',
+  '$SWARM_ID',
+  json('{"iterations": $ITERATIONS, "final_confidence": $CONFIDENCE}'),
+  json('{"iterate_feedback": $(echo "$FEEDBACK" | jq -Rs .)}'),
+  json('$EXTRACTED_LESSONS'),
+  json('$METADATA'),
+  $CONFIDENCE,
+  datetime('now')
+);
+EOF
+
+  # Output result
+  RESULT="{
+  \"id\": \"$REFLECTION_ID\",
+  \"reflection_type\": \"$LESSON_TYPE\",
+  \"task_id\": \"$TASK_ID\",
+  \"severity\": \"$SEVERITY\",
+  \"confidence\": $CONFIDENCE,
+  \"iterations\": $ITERATIONS,
+  \"failure_reason\": \"$FAILURE_REASON\",
+  \"solution\": \"$SOLUTION\",
+  \"stored\": true
+}"
+
+  if [ -n "$OUTPUT" ]; then
+    echo "$RESULT" > "$OUTPUT"
+    echo "Anti-pattern reflection saved to: $OUTPUT" >&2
+  else
+    echo "$RESULT"
+  fi
+
+  exit 0
+fi
+
+# Legacy Mode - Original ACE reflector
+if [ -z "$CONTEXT" ]; then
+  echo "Error: Either --context (legacy mode) or --confidence + --iterations + --feedback + --task-id (anti-pattern mode) required"
+  echo ""
+  echo "Legacy Mode Usage:"
+  echo "  $0 --context '{\"task\": \"...\"}' [OPTIONS]"
+  echo ""
+  echo "Anti-Pattern Mode Usage:"
+  echo "  $0 --confidence 0.45 --iterations 3 --feedback \"Missing error handling\" --task-id \"sprint-001\" [OPTIONS]"
+  exit 1
+fi
+
+# Validate JSON
+if ! echo "$CONTEXT" | jq . > /dev/null 2>&1; then
+  echo "Error: Invalid JSON context"
+  exit 1
+fi
+
+# Execute reflection using Node.js with inline code (original implementation)
+if [ -n "$OUTPUT" ]; then
+  node --input-type=module --eval "
+import ACEReflector from '${PROJECT_ROOT}/dist/ace/ace-reflector.js';
+
+const context = JSON.parse(process.argv[1]);
+const complexity = process.argv[2] ? parseFloat(process.argv[2]) : undefined;
+const memoryPath = process.argv[3];
+
+const reflector = new ACEReflector(memoryPath);
+await reflector.initialize();
+
+const reflection = await reflector.reflect(context, { complexity });
+console.log(JSON.stringify(reflection, null, 2));
+" "$CONTEXT" "$COMPLEXITY" "$MEMORY_PATH" > "$OUTPUT"
+  echo "Reflection saved to: $OUTPUT" >&2
+else
+  node --input-type=module --eval "
+import ACEReflector from '${PROJECT_ROOT}/dist/ace/ace-reflector.js';
+
+const context = JSON.parse(process.argv[1]);
+const complexity = process.argv[2] ? parseFloat(process.argv[2]) : undefined;
+const memoryPath = process.argv[3];
+
+const reflector = new ACEReflector(memoryPath);
+await reflector.initialize();
+
+const reflection = await reflector.reflect(context, { complexity });
+console.log(JSON.stringify(reflection, null, 2));
+" "$CONTEXT" "$COMPLEXITY" "$MEMORY_PATH"
+fi

package/claude-assets/skills/cfn-ace-system/invoke-context-stats.sh

@@ -0,0 +1,227 @@
+#!/usr/bin/env bash
+
+##############################################################################
+# ACE Context Statistics Wrapper
+# Queries reflection database for statistics and analysis
+#
+# Usage:
+#   ./invoke-context-stats.sh --query <type> [OPTIONS]
+#
+# Arguments:
+#   --query            Query type: reflections, insights, summary (required)
+#   --filter           MongoDB-style filter JSON (optional)
+#   --limit            Maximum results (default: 100)
+#   --memory-path      SQLite memory path (optional)
+##############################################################################
+
+set -euo pipefail
+
+# Default values
+QUERY=""
+FILTER="{}"
+LIMIT=100
+MEMORY_PATH="${ACE_MEMORY_PATH:-./.artifacts/database/swarm-memory.db}"
+
+# Parse arguments
+while [[ $# -gt 0 ]]; do
+  case $1 in
+    --query)
+      QUERY="$2"
+      shift 2
+      ;;
+    --filter)
+      FILTER="$2"
+      shift 2
+      ;;
+    --limit)
+      LIMIT="$2"
+      shift 2
+      ;;
+    --memory-path)
+      MEMORY_PATH="$2"
+      shift 2
+      ;;
+    --show-indexes)
+      QUERY="indexes"
+      shift
+      ;;
+    *)
+      echo "Unknown option: $1"
+      exit 1
+      ;;
+  esac
+done
+
+# Validation
+if [ -z "$QUERY" ]; then
+  echo "Error: --query is required"
+  echo "Usage: $0 --query <reflections|insights|summary> [OPTIONS]"
+  exit 1
+fi
+
+# Validate filter JSON
+if ! echo "$FILTER" | jq . > /dev/null 2>&1; then
+  echo "Error: Invalid JSON filter"
+  exit 1
+fi
+
+# Get project root
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+PROJECT_ROOT="$(cd "$SCRIPT_DIR/../../.." && pwd)"
+
+# Create Node.js runner script
+RUNNER_SCRIPT=$(mktemp --suffix=.mjs)
+trap "rm -f $RUNNER_SCRIPT" EXIT
+
+cat > "$RUNNER_SCRIPT" << EOF
+import { SQLiteMemorySystem } from '${PROJECT_ROOT}/dist/memory/sqlite-memory-system.js';
+import { AccessLevel } from '${PROJECT_ROOT}/dist/memory/memory-adapter.js';
+
+const queryType = process.argv[2];
+const filter = JSON.parse(process.argv[3]);
+const limit = parseInt(process.argv[4]);
+const memoryPath = process.argv[5];
+
+const memorySystem = new SQLiteMemorySystem(memoryPath);
+await memorySystem.initialize();
+
+async function queryReflections() {
+  // Query all reflections matching filter
+  const db = memorySystem['db'];
+  if (!db) {
+    throw new Error('Database not initialized');
+  }
+
+  let sql = 'SELECT * FROM context_reflections';
+  const whereClauses = [];
+  const params = [];
+
+  // Apply filters
+  if (filter.complexity) {
+    if (filter.complexity['\$gt'] !== undefined) {
+      whereClauses.push('complexity > ?');
+      params.push(filter.complexity['\$gt']);
+    }
+    if (filter.complexity['\$lt'] !== undefined) {
+      whereClauses.push('complexity < ?');
+      params.push(filter.complexity['\$lt']);
+    }
+  }
+
+  if (filter.timestamp) {
+    if (filter.timestamp['\$gt'] !== undefined) {
+      whereClauses.push('timestamp > ?');
+      params.push(filter.timestamp['\$gt']);
+    }
+  }
+
+  if (whereClauses.length > 0) {
+    sql += ' WHERE ' + whereClauses.join(' AND ');
+  }
+
+  sql += ' ORDER BY timestamp DESC LIMIT ?';
+  params.push(limit);
+
+  const rows = db.prepare(sql).all(...params);
+
+  return rows.map(row => ({
+    id: row.id,
+    timestamp: row.timestamp,
+    complexity: row.complexity,
+    context: JSON.parse(row.context),
+    insights: JSON.parse(row.insights)
+  }));
+}
+
+async function queryInsights() {
+  const reflections = await queryReflections();
+  const allInsights = reflections.flatMap(r => r.insights);
+
+  // Count insight frequency
+  const insightCounts = {};
+  allInsights.forEach(insight => {
+    insightCounts[insight] = (insightCounts[insight] || 0) + 1;
+  });
+
+  return Object.entries(insightCounts)
+    .map(([insight, count]) => ({ insight, count }))
+    .sort((a, b) => b.count - a.count);
+}
+
+async function querySummary() {
+  const reflections = await queryReflections();
+
+  const complexities = reflections.map(r => r.complexity);
+  const avgComplexity = complexities.reduce((a, b) => a + b, 0) / complexities.length;
+  const maxComplexity = Math.max(...complexities);
+  const minComplexity = Math.min(...complexities);
+
+  return {
+    totalReflections: reflections.length,
+    avgComplexity: avgComplexity || 0,
+    maxComplexity: maxComplexity || 0,
+    minComplexity: minComplexity || 0,
+    timeRange: {
+      earliest: Math.min(...reflections.map(r => r.timestamp)),
+      latest: Math.max(...reflections.map(r => r.timestamp))
+    }
+  };
+}
+
+async function queryIndexes() {
+  const db = memorySystem['db'];
+  if (!db) {
+    throw new Error('Database not initialized');
+  }
+
+  // Get all indexes on context_reflections table
+  const indexes = await db.all(
+    "SELECT name, sql FROM sqlite_master WHERE type='index' AND tbl_name='context_reflections' AND name NOT LIKE 'sqlite_%'"
+  );
+
+  // Get index usage stats from last query (if available)
+  const totalIndexes = indexes.length;
+
+  return {
+    totalIndexes,
+    indexes: indexes.map(idx => ({
+      name: idx.name,
+      definition: idx.sql || 'auto-index'
+    })),
+    status: totalIndexes >= 6 ? 'optimized' : 'partial',
+    expectedIndexes: [
+      'idx_reflections_tags',
+      'idx_reflections_domain',
+      'idx_reflections_confidence',
+      'idx_reflections_created_at',
+      'idx_reflections_domain_conf_date',
+      'idx_reflections_conf_date'
+    ]
+  };
+}
+
+let result;
+
+switch (queryType) {
+  case 'reflections':
+    result = await queryReflections();
+    break;
+  case 'insights':
+    result = await queryInsights();
+    break;
+  case 'summary':
+    result = await querySummary();
+    break;
+  case 'indexes':
+    result = await queryIndexes();
+    break;
+  default:
+    throw new Error(`Unknown query type: ${queryType}`);
+}
+
+console.log(JSON.stringify(result, null, 2));
+EOF
+
+# Execute query
+cd "$PROJECT_ROOT"
+node "$RUNNER_SCRIPT" "$QUERY" "$FILTER" "$LIMIT" "$MEMORY_PATH"

package/claude-assets/skills/cfn-ace-system/log-merge.sh

@@ -0,0 +1,67 @@
+#!/bin/bash
+# Log merge action to context_merge_log
+# Usage: ./log-merge.sh --merge-type new_bullet --bullet-id STRAT-007 --reflection-id refl-123
+
+set -e
+
+DB_PATH="${ACE_DB_PATH:-./.artifacts/database/swarm-memory.db}"
+MERGE_TYPE=""
+BULLET_ID=""
+REFLECTION_ID=""
+SIMILARITY_SCORE=""
+CURATOR_REASONING=""
+MERGED_FROM_IDS="[]"
+
+while [[ $# -gt 0 ]]; do
+  case $1 in
+    --merge-type) MERGE_TYPE="$2"; shift 2 ;;
+    --bullet-id) BULLET_ID="$2"; shift 2 ;;
+    --reflection-id) REFLECTION_ID="$2"; shift 2 ;;
+    --similarity-score) SIMILARITY_SCORE="$2"; shift 2 ;;
+    --curator-reasoning) CURATOR_REASONING="$2"; shift 2 ;;
+    --merged-from-ids) MERGED_FROM_IDS="$2"; shift 2 ;;
+    *) echo "Unknown option: $1"; exit 1 ;;
+  esac
+done
+
+if [ -z "$MERGE_TYPE" ] || [ -z "$BULLET_ID" ]; then
+  echo "Error: --merge-type and --bullet-id required"
+  exit 1
+fi
+
+LOG_ID="merge-$(date +%s)-$(openssl rand -hex 3)"
+
+sqlite3 "$DB_PATH" <<EOF
+CREATE TABLE IF NOT EXISTS context_merge_log (
+    id TEXT PRIMARY KEY,
+    merge_type TEXT NOT NULL CHECK (merge_type IN ('new_bullet', 'increment_helpful', 'increment_harmful', 'merge_similar', 'archive', 'edit', 'version_bump')),
+    bullet_id TEXT NOT NULL,
+    reflection_id TEXT,
+    old_content TEXT,
+    new_content TEXT,
+    similarity_score REAL,
+    merged_from_bullet_ids TEXT,
+    curator_agent_id TEXT,
+    curator_reasoning TEXT,
+    acl_level INTEGER NOT NULL DEFAULT 5 CHECK (acl_level BETWEEN 1 AND 6),
+    swarm_id TEXT,
+    project_id TEXT,
+    created_at DATETIME DEFAULT CURRENT_TIMESTAMP
+);
+
+INSERT INTO context_merge_log (
+  id, merge_type, bullet_id, reflection_id, similarity_score,
+  merged_from_bullet_ids, curator_reasoning, acl_level
+) VALUES (
+  '${LOG_ID}',
+  '${MERGE_TYPE}',
+  '${BULLET_ID}',
+  $([ -n "$REFLECTION_ID" ] && echo "'$REFLECTION_ID'" || echo "NULL"),
+  $([ -n "$SIMILARITY_SCORE" ] && echo "${SIMILARITY_SCORE}" || echo "NULL"),
+  '${MERGED_FROM_IDS}',
+  '${CURATOR_REASONING}',
+  5
+);
+EOF
+
+echo "{\"status\":\"success\",\"log_id\":\"${LOG_ID}\",\"merge_type\":\"${MERGE_TYPE}\"}"