agentsys 5.0.0

package/plugins/enhance/lib/enhance/projectmemory-analyzer.js

@@ -0,0 +1,551 @@
+/**
+ * Project Memory Analyzer
+ * Analyzes CLAUDE.md/AGENTS.md project memory files for optimization opportunities
+ */
+
+const fs = require('fs');
+const path = require('path');
+const { projectMemoryPatterns } = require('./projectmemory-patterns');
+
+const PROJECT_MEMORY_FILES = [
+  'CLAUDE.md',
+  'AGENTS.md',
+  '.github/CLAUDE.md',
+  '.github/AGENTS.md'
+];
+
+/**
+ * Find the project memory file in a directory
+ * @param {string} projectPath - Project root directory
+ * @returns {Object|null} { path, name, type } or null if not found
+ */
+function findProjectMemoryFile(projectPath) {
+  for (const fileName of PROJECT_MEMORY_FILES) {
+    const filePath = path.join(projectPath, fileName);
+    if (fs.existsSync(filePath)) {
+      return {
+        path: filePath,
+        name: fileName,
+        type: fileName.includes('AGENTS') ? 'agents' : 'claude'
+      };
+    }
+  }
+  return null;
+}
+
+/**
+ * Extract file references from markdown content
+ * @param {string} content - Markdown content
+ * @returns {Array} Array of file paths referenced
+ */
+function extractFileReferences(content) {
+  if (!content || typeof content !== 'string') return [];
+
+  const references = [];
+
+  // Extract markdown links [text](path) using indexOf scanning (ReDoS-safe)
+  let pos = 0;
+  while (pos < content.length) {
+    const openBracket = content.indexOf('[', pos);
+    if (openBracket === -1) break;
+    const closeBracket = content.indexOf(']', openBracket + 1);
+    if (closeBracket === -1) break;
+    if (content[closeBracket + 1] === '(') {
+      const closeParen = content.indexOf(')', closeBracket + 2);
+      if (closeParen !== -1 && closeParen - closeBracket - 2 <= 500) {
+        const href = content.substring(closeBracket + 2, closeParen);
+        if (!href.startsWith('http') && !href.startsWith('#') && !href.startsWith('mailto:')) {
+          references.push(href.split('#')[0]); // Remove anchor
+        }
+        pos = closeParen + 1;
+        continue;
+      }
+    }
+    pos = openBracket + 1;
+  }
+
+  // Match backtick paths: `path/to/file.ext` or `file.ext` (root files)
+  const backtickMatches = content.match(/`([^`]+)`/g) || [];
+  for (const match of backtickMatches) {
+    const filePath = match.replace(/`/g, '');
+    // Include paths with / or extension, exclude spaces and variables
+    if ((filePath.includes('.') || filePath.includes('/')) && !filePath.includes(' ') && !filePath.startsWith('$')) {
+      references.push(filePath);
+    }
+  }
+
+  return [...new Set(references)];
+}
+
+/**
+ * Validate file references exist
+ * @param {string} content - Markdown content
+ * @param {string} projectPath - Project root directory
+ * @returns {Object} { valid: [], broken: [] }
+ */
+function validateFileReferences(content, projectPath) {
+  const references = extractFileReferences(content);
+  const valid = [];
+  const broken = [];
+
+  const resolvedProjectPath = path.resolve(projectPath);
+
+  for (const ref of references) {
+    // Skip glob patterns and variable references
+    if (ref.includes('*') || ref.includes('${')) {
+      valid.push(ref);
+      continue;
+    }
+
+    // Resolve path and validate it stays within project root (prevent path traversal)
+    // Use path.relative() for Windows case-insensitive path comparison
+    const fullPath = path.resolve(projectPath, ref);
+    const relativePath = path.relative(resolvedProjectPath, fullPath);
+    if (!relativePath.startsWith('..') && !path.isAbsolute(relativePath) && fs.existsSync(fullPath)) {
+      valid.push(ref);
+    } else {
+      broken.push(ref);
+    }
+  }
+
+  return { valid, broken };
+}
+
+/**
+ * Extract npm script references from content
+ * @param {string} content - Markdown content
+ * @returns {Array} Array of npm commands referenced
+ */
+function extractCommandReferences(content) {
+  if (!content || typeof content !== 'string') return [];
+
+  const commands = [];
+
+  // Match npm run commands: npm run <script>
+  const npmRunMatches = content.match(/npm\s+run\s+([a-z][\w:-]*)/gi) || [];
+  for (const match of npmRunMatches) {
+    const scriptMatch = match.match(/npm\s+run\s+([a-z][\w:-]*)/i);
+    if (scriptMatch && scriptMatch[1]) {
+      commands.push(scriptMatch[1]);
+    }
+  }
+
+  // Match npm <script> shorthand (test, start, etc)
+  const npmShortMatches = content.match(/npm\s+(test|start|build|lint|install)/gi) || [];
+  for (const match of npmShortMatches) {
+    const cmdMatch = match.match(/npm\s+(\w+)/i);
+    if (cmdMatch && cmdMatch[1]) {
+      commands.push(cmdMatch[1]);
+    }
+  }
+
+  return [...new Set(commands)];
+}
+
+/**
+ * Validate command references exist in package.json
+ * @param {string} content - Markdown content
+ * @param {string} projectPath - Project root directory
+ * @returns {Object} { valid: [], broken: [] }
+ */
+function validateCommandReferences(content, projectPath) {
+  const commands = extractCommandReferences(content);
+  const valid = [];
+  const broken = [];
+
+  const packagePath = path.join(projectPath, 'package.json');
+  let scripts = {};
+
+  let packageParseError = null;
+  if (fs.existsSync(packagePath)) {
+    try {
+      const pkg = JSON.parse(fs.readFileSync(packagePath, 'utf8'));
+      scripts = pkg.scripts || {};
+    } catch (err) {
+      packageParseError = err.message;
+      // Continue with empty scripts - commands will be marked as broken
+    }
+  }
+
+  const builtInCommands = ['install', 'i', 'ci', 'test', 't', 'start', 'build', 'publish', 'pack'];
+
+  for (const cmd of commands) {
+    if (scripts[cmd] || builtInCommands.includes(cmd)) {
+      valid.push(cmd);
+    } else {
+      broken.push(cmd);
+    }
+  }
+
+  const result = { valid, broken };
+  if (packageParseError) {
+    result.parseError = packageParseError;
+  }
+  return result;
+}
+
+/**
+ * Calculate token metrics for content
+ * @param {string} content - Content to analyze
+ * @param {string} readmeContent - Optional README content for comparison
+ * @returns {Object} Token metrics
+ */
+function calculateTokenMetrics(content, readmeContent = null) {
+  if (!content || typeof content !== 'string') {
+    return { estimatedTokens: 0, characterCount: 0, lineCount: 0, wordCount: 0 };
+  }
+
+  const characterCount = content.length;
+  const lineCount = content.split('\n').length;
+  const estimatedTokens = Math.ceil(characterCount / 4);
+
+  const result = {
+    estimatedTokens,
+    characterCount,
+    lineCount,
+    wordCount: content.split(/\s+/).filter(Boolean).length
+  };
+
+  // Calculate README overlap if provided
+  if (readmeContent && typeof readmeContent === 'string') {
+    result.readmeOverlap = calculateTextOverlap(content, readmeContent);
+  }
+
+  return result;
+}
+
+/**
+ * Calculate text overlap between two documents
+ * @param {string} text1 - First text
+ * @param {string} text2 - Second text
+ * @returns {number} Overlap ratio (0-1)
+ */
+function calculateTextOverlap(text1, text2) {
+  if (!text1 || !text2) return 0;
+
+  const normalize = (text) => text.toLowerCase().replace(/[^\w\s]/g, ' ').replace(/\s+/g, ' ').trim();
+  const sentences1 = normalize(text1).split(/[.!?]+/).filter(s => s.length > 20);
+  const sentences2 = new Set(normalize(text2).split(/[.!?]+/).filter(s => s.length > 20));
+
+  if (sentences1.length === 0) return 0;
+
+  let matchCount = 0;
+  for (const sentence of sentences1) {
+    for (const s2 of sentences2) {
+      const words1 = sentence.split(' ');
+      const words2 = new Set(s2.split(' '));
+      const matchingWords = words1.filter(w => words2.has(w)).length;
+      if (matchingWords / words1.length > 0.8) {
+        matchCount++;
+        break;
+      }
+    }
+  }
+
+  return matchCount / sentences1.length;
+}
+
+/**
+ * Detect README duplication
+ * @param {string} memoryFilePath - Path to project memory file
+ * @param {string} readmePath - Path to README.md (optional, auto-detected)
+ * @returns {Object} Duplication analysis
+ */
+function detectReadmeDuplication(memoryFilePath, readmePath = null) {
+  if (!readmePath) {
+    const possibleReadmes = ['README.md', 'readme.md', 'Readme.md'];
+    let currentDir = path.dirname(memoryFilePath);
+
+    // Walk up directory tree to find README (handles .github/ case)
+    while (currentDir && currentDir.length > 0) {
+      for (const name of possibleReadmes) {
+        const tryPath = path.join(currentDir, name);
+        if (fs.existsSync(tryPath)) {
+          readmePath = tryPath;
+          break;
+        }
+      }
+      if (readmePath) break;
+
+      const parentDir = path.dirname(currentDir);
+      if (!parentDir || parentDir === currentDir) break;
+      currentDir = parentDir;
+    }
+  }
+
+  if (!readmePath || !fs.existsSync(readmePath)) {
+    return { hasReadme: false, duplicationRatio: 0 };
+  }
+
+  try {
+    const memoryContent = fs.readFileSync(memoryFilePath, 'utf8');
+    const readmeContent = fs.readFileSync(readmePath, 'utf8');
+
+    return {
+      hasReadme: true,
+      readmePath,
+      duplicationRatio: calculateTextOverlap(memoryContent, readmeContent)
+    };
+  } catch (err) {
+    return { hasReadme: true, readmePath, error: err.message };
+  }
+}
+
+/**
+ * Analyze a single project memory file
+ * @param {string} filePath - Path to project memory file
+ * @param {Object} options - Analysis options
+ * @param {boolean} options.verbose - Include LOW certainty issues
+ * @param {boolean} options.checkReferences - Validate file/command references
+ * @returns {Object} Analysis results
+ */
+function analyzeFile(filePath, options = {}) {
+  const { verbose = false, checkReferences = true } = options;
+
+  const results = {
+    fileName: path.basename(filePath),
+    filePath,
+    fileType: filePath.includes('AGENTS') ? 'agents' : 'claude',
+    structureIssues: [],
+    referenceIssues: [],
+    efficiencyIssues: [],
+    qualityIssues: [],
+    crossPlatformIssues: [],
+    metrics: null
+  };
+
+  // Read file
+  if (!fs.existsSync(filePath)) {
+    results.structureIssues.push({
+      issue: 'File not found',
+      file: filePath,
+      certainty: 'HIGH',
+      patternId: 'file_not_found'
+    });
+    return results;
+  }
+
+  let content;
+  try {
+    content = fs.readFileSync(filePath, 'utf8');
+  } catch (err) {
+    results.structureIssues.push({
+      issue: `Failed to read file: ${err.message}`,
+      file: filePath,
+      certainty: 'HIGH',
+      patternId: 'read_error'
+    });
+    return results;
+  }
+
+  // Determine project root - if file is in .github/, use parent directory
+  let projectPath = path.dirname(filePath);
+  if (projectPath.endsWith('.github') || projectPath.endsWith('.github/') || projectPath.endsWith('.github\\')) {
+    projectPath = path.dirname(projectPath);
+  }
+
+  const readmeDuplication = detectReadmeDuplication(filePath);
+  let readmeContent = null;
+  if (readmeDuplication.hasReadme && !readmeDuplication.error) {
+    try {
+      readmeContent = fs.readFileSync(readmeDuplication.readmePath, 'utf8');
+    } catch (err) {
+      // README exists but couldn't be read - continue without it
+    }
+  }
+  results.metrics = calculateTokenMetrics(content, readmeContent);
+
+  const context = {
+    fileName: results.fileName,
+    duplicationRatio: readmeDuplication.duplicationRatio
+  };
+
+  if (checkReferences) {
+    const fileRefs = validateFileReferences(content, projectPath);
+    context.brokenFiles = fileRefs.broken;
+
+    const cmdRefs = validateCommandReferences(content, projectPath);
+    context.brokenCommands = cmdRefs.broken;
+  }
+
+  const structurePatterns = ['missing_critical_rules', 'missing_architecture', 'missing_key_commands'];
+  for (const patternName of structurePatterns) {
+    const pattern = projectMemoryPatterns[patternName];
+    if (!pattern) continue;
+
+    const result = pattern.check(content, context);
+    if (result && (verbose || pattern.certainty !== 'LOW')) {
+      results.structureIssues.push({
+        ...result,
+        file: filePath,
+        certainty: pattern.certainty,
+        patternId: pattern.id
+      });
+    }
+  }
+
+  const referencePatterns = ['broken_file_reference', 'broken_command_reference'];
+  for (const patternName of referencePatterns) {
+    const pattern = projectMemoryPatterns[patternName];
+    if (!pattern) continue;
+
+    const result = pattern.check(content, context);
+    if (result && (verbose || pattern.certainty !== 'LOW')) {
+      results.referenceIssues.push({
+        ...result,
+        file: filePath,
+        certainty: pattern.certainty,
+        patternId: pattern.id
+      });
+    }
+  }
+
+  const efficiencyPatterns = ['readme_duplication', 'excessive_token_count', 'verbose_instructions', 'example_overload'];
+  for (const patternName of efficiencyPatterns) {
+    const pattern = projectMemoryPatterns[patternName];
+    if (!pattern) continue;
+
+    const result = pattern.check(content, context);
+    if (result && (verbose || pattern.certainty !== 'LOW')) {
+      results.efficiencyIssues.push({
+        ...result,
+        file: filePath,
+        certainty: pattern.certainty,
+        patternId: pattern.id
+      });
+    }
+  }
+
+  const qualityPatterns = ['missing_why', 'deep_nesting'];
+  for (const patternName of qualityPatterns) {
+    const pattern = projectMemoryPatterns[patternName];
+    if (!pattern) continue;
+
+    const result = pattern.check(content, context);
+    if (result && (verbose || pattern.certainty !== 'LOW')) {
+      results.qualityIssues.push({
+        ...result,
+        file: filePath,
+        certainty: pattern.certainty,
+        patternId: pattern.id
+      });
+    }
+  }
+
+  const crossPlatformPatterns = ['hardcoded_state_dir', 'claude_only_terminology', 'missing_agents_md_mention'];
+  for (const patternName of crossPlatformPatterns) {
+    const pattern = projectMemoryPatterns[patternName];
+    if (!pattern) continue;
+
+    const result = pattern.check(content, context);
+    if (result && (verbose || pattern.certainty !== 'LOW')) {
+      results.crossPlatformIssues.push({
+        ...result,
+        file: filePath,
+        certainty: pattern.certainty,
+        patternId: pattern.id
+      });
+    }
+  }
+
+  return results;
+}
+
+/**
+ * Main analyze function
+ * @param {string} targetPath - Path to project directory or specific file
+ * @param {Object} options - Analysis options
+ * @param {boolean} options.verbose - Include LOW certainty issues
+ * @param {boolean} options.checkReferences - Validate file/command references
+ * @returns {Object} Analysis results
+ */
+function analyze(targetPath, options = {}) {
+  let filePath;
+
+  if (fs.existsSync(targetPath)) {
+    const stat = fs.statSync(targetPath);
+    if (stat.isFile()) {
+      filePath = targetPath;
+    } else {
+      const found = findProjectMemoryFile(targetPath);
+      if (!found) {
+        return {
+          error: 'No project memory file found',
+          searchedPaths: PROJECT_MEMORY_FILES.map(f => path.join(targetPath, f)),
+          structureIssues: [{
+            issue: 'No CLAUDE.md or AGENTS.md found in project',
+            fix: 'Create CLAUDE.md with project memory content',
+            certainty: 'HIGH',
+            patternId: 'missing_project_memory'
+          }]
+        };
+      }
+      filePath = found.path;
+    }
+  } else {
+    return {
+      error: `Path does not exist: ${targetPath}`,
+      structureIssues: [{
+        issue: 'Target path does not exist',
+        file: targetPath,
+        certainty: 'HIGH',
+        patternId: 'path_not_found'
+      }]
+    };
+  }
+
+  return analyzeFile(filePath, options);
+}
+
+/**
+ * Apply fixes to analysis results
+ * @param {Object} results - Analysis results
+ * @param {Object} options - Fix options
+ * @returns {Object} Fix results
+ */
+function applyFixes(results, options = {}) {
+  const allIssues = [
+    ...(results.structureIssues || []),
+    ...(results.referenceIssues || []),
+    ...(results.efficiencyIssues || []),
+    ...(results.qualityIssues || []),
+    ...(results.crossPlatformIssues || [])
+  ];
+
+  return {
+    applied: [],
+    skipped: allIssues.map(i => ({
+      ...i,
+      reason: 'No auto-fix available - requires human judgment'
+    })),
+    errors: []
+  };
+}
+
+/**
+ * Generate markdown report from analysis results
+ * @param {Object} results - Analysis results
+ * @param {Object} options - Report options
+ * @returns {string} Markdown report
+ */
+function generateReport(results, options = {}) {
+  const reporter = require('./reporter');
+
+  return reporter.generateProjectMemoryReport(results, options);
+}
+
+module.exports = {
+  PROJECT_MEMORY_FILES,
+  findProjectMemoryFile,
+  extractFileReferences,
+  validateFileReferences,
+  extractCommandReferences,
+  validateCommandReferences,
+  calculateTokenMetrics,
+  calculateTextOverlap,
+  detectReadmeDuplication,
+  analyzeFile,
+  analyze,
+  applyFixes,
+  generateReport
+};