@agenticmail/enterprise 0.3.2 → 0.4.0

package/src/agent-tools/tools/enterprise-logs.ts

@@ -0,0 +1,479 @@
+/**
+ * AgenticMail Agent Tools — Enterprise Log Analysis
+ *
+ * File-based log search, aggregation, tailing, correlation, and error extraction.
+ * Uses fs.createReadStream + readline for efficient large file handling.
+ */
+
+import fs from 'node:fs';
+import fsPromises from 'node:fs/promises';
+import path from 'node:path';
+import readline from 'node:readline';
+import type { AnyAgentTool, ToolCreationOptions } from '../types.js';
+import { readStringParam, readNumberParam, jsonResult, textResult, errorResult } from '../common.js';
+
+var DEFAULT_SEARCH_LIMIT = 100;
+var DEFAULT_TAIL_LINES = 50;
+var DEFAULT_ERROR_LIMIT = 50;
+var DEFAULT_CORRELATE_LIMIT = 50;
+var ERROR_PATTERNS = /\b(ERROR|FATAL|CRITICAL|Exception|Traceback|panic|PANIC)\b/i;
+var TIMESTAMP_PATTERN = /\d{4}-\d{2}-\d{2}[T ]\d{2}:\d{2}:\d{2}/;
+
+// --- File collection ---
+
+async function collectLogFiles(targetPath: string): Promise<string[]> {
+  var results: string[] = [];
+
+  try {
+    var stat = await fsPromises.stat(targetPath);
+    if (stat.isFile()) return [targetPath];
+  } catch {
+    return [];
+  }
+
+  async function walk(dir: string) {
+    try {
+      var entries = await fsPromises.readdir(dir, { withFileTypes: true });
+      for (var entry of entries) {
+        if (entry.name.startsWith('.')) continue;
+        var full = path.join(dir, entry.name);
+        if (entry.isDirectory()) {
+          await walk(full);
+        } else {
+          var ext = path.extname(entry.name).toLowerCase();
+          if (ext === '.log' || ext === '.txt' || ext === '' || ext === '.json' || ext === '.out' || ext === '.err') {
+            results.push(full);
+          }
+        }
+      }
+    } catch { /* skip unreadable dirs */ }
+  }
+
+  await walk(targetPath);
+  return results.sort();
+}
+
+// --- Line-by-line reader ---
+
+async function readLines(
+  filePath: string,
+  callback: (line: string, lineNum: number, file: string) => boolean | void
+): Promise<void> {
+  return new Promise(function(resolve, reject) {
+    var stream = fs.createReadStream(filePath, { encoding: 'utf-8' });
+    var rl = readline.createInterface({ input: stream, crlfDelay: Infinity });
+    var lineNum = 0;
+    var stopped = false;
+
+    rl.on('line', function(line) {
+      lineNum++;
+      if (stopped) return;
+      var result = callback(line, lineNum, filePath);
+      if (result === false) {
+        stopped = true;
+        rl.close();
+        stream.destroy();
+      }
+    });
+
+    rl.on('close', function() { resolve(); });
+    rl.on('error', function(err) { reject(err); });
+    stream.on('error', function(err) { reject(err); });
+  });
+}
+
+// --- Timestamp parsing ---
+
+function extractTimestamp(line: string): string | null {
+  var match = line.match(TIMESTAMP_PATTERN);
+  return match ? match[0] : null;
+}
+
+function isWithinTimeRange(timestamp: string | null, since?: string, until?: string): boolean {
+  if (!timestamp) return true; // Include lines without timestamps
+  if (since && timestamp < since) return false;
+  if (until && timestamp > until) return false;
+  return true;
+}
+
+// --- Resolve path ---
+
+function resolvePath(targetPath: string, options?: ToolCreationOptions): string {
+  if (path.isAbsolute(targetPath)) return targetPath;
+  var workDir = options?.workspaceDir || process.cwd();
+  return path.resolve(workDir, targetPath);
+}
+
+export function createEnterpriseLogTools(options?: ToolCreationOptions): AnyAgentTool[] {
+  return [
+    {
+      name: 'ent_log_search',
+      label: 'Search Logs',
+      description: 'Search log files for a pattern (string or regex). Supports time range filtering. Returns matching lines with file path and line number.',
+      category: 'search',
+      risk: 'low',
+      parameters: {
+        type: 'object',
+        properties: {
+          path: { type: 'string', description: 'Log file or directory path.' },
+          pattern: { type: 'string', description: 'Search pattern (string or regex).' },
+          since: { type: 'string', description: 'Start time filter (ISO date, e.g. "2024-01-15T10:00:00").' },
+          until: { type: 'string', description: 'End time filter (ISO date).' },
+          limit: { type: 'number', description: 'Maximum matching lines to return (default 100).' },
+        },
+        required: ['path', 'pattern'],
+      },
+      execute: async function(_toolCallId, args) {
+        var params = args as Record<string, unknown>;
+        var targetPath = resolvePath(readStringParam(params, 'path', { required: true }), options);
+        var pattern = readStringParam(params, 'pattern', { required: true });
+        var since = readStringParam(params, 'since');
+        var until = readStringParam(params, 'until');
+        var limit = readNumberParam(params, 'limit', { integer: true }) ?? DEFAULT_SEARCH_LIMIT;
+
+        var regex: RegExp;
+        try {
+          regex = new RegExp(pattern, 'i');
+        } catch {
+          // Treat as literal string if invalid regex
+          regex = new RegExp(pattern.replace(/[.*+?^${}()|[\]\\]/g, '\\$&'), 'i');
+        }
+
+        var files = await collectLogFiles(targetPath);
+        if (files.length === 0) {
+          return errorResult('No log files found at: ' + targetPath);
+        }
+
+        var matches: Array<{ file: string; line: number; text: string; timestamp: string | null }> = [];
+
+        for (var file of files) {
+          if (matches.length >= limit) break;
+          try {
+            await readLines(file, function(line, lineNum, filePath) {
+              if (matches.length >= limit) return false;
+              var timestamp = extractTimestamp(line);
+              if (!isWithinTimeRange(timestamp, since, until)) return;
+              if (regex.test(line)) {
+                matches.push({
+                  file: path.relative(resolvePath('.', options), filePath),
+                  line: lineNum,
+                  text: line.length > 500 ? line.slice(0, 500) + '...' : line,
+                  timestamp: timestamp,
+                });
+              }
+            });
+          } catch { /* skip unreadable files */ }
+        }
+
+        if (matches.length === 0) {
+          return textResult('No matches found for "' + pattern + '" in ' + files.length + ' file(s).');
+        }
+
+        return jsonResult({
+          pattern: pattern,
+          filesSearched: files.length,
+          matchCount: matches.length,
+          truncated: matches.length >= limit,
+          matches: matches,
+        });
+      },
+    },
+
+    {
+      name: 'ent_log_aggregate',
+      label: 'Aggregate Logs',
+      description: 'Count log entries grouped by a field (e.g. "level", "status_code", or a regex capture group). Returns sorted counts.',
+      category: 'search',
+      risk: 'low',
+      parameters: {
+        type: 'object',
+        properties: {
+          path: { type: 'string', description: 'Log file or directory path.' },
+          group_by: { type: 'string', description: 'Grouping field: "level", "status_code", or a regex with a capture group.' },
+          since: { type: 'string', description: 'Start time filter (ISO date).' },
+          until: { type: 'string', description: 'End time filter (ISO date).' },
+        },
+        required: ['path', 'group_by'],
+      },
+      execute: async function(_toolCallId, args) {
+        var params = args as Record<string, unknown>;
+        var targetPath = resolvePath(readStringParam(params, 'path', { required: true }), options);
+        var groupBy = readStringParam(params, 'group_by', { required: true });
+        var since = readStringParam(params, 'since');
+        var until = readStringParam(params, 'until');
+
+        var files = await collectLogFiles(targetPath);
+        if (files.length === 0) {
+          return errorResult('No log files found at: ' + targetPath);
+        }
+
+        // Build extraction function based on group_by
+        var extractGroup: (line: string) => string | null;
+
+        if (groupBy === 'level') {
+          extractGroup = function(line) {
+            var match = line.match(/\b(DEBUG|INFO|WARN|WARNING|ERROR|FATAL|CRITICAL|TRACE)\b/i);
+            return match ? match[1].toUpperCase() : null;
+          };
+        } else if (groupBy === 'status_code') {
+          extractGroup = function(line) {
+            var match = line.match(/\b(status|code|HTTP\/\d\.\d"?\s+)(\d{3})\b/i);
+            return match ? match[2] : null;
+          };
+        } else {
+          // Try as regex with capture group
+          var customRegex: RegExp;
+          try {
+            customRegex = new RegExp(groupBy, 'i');
+          } catch {
+            return errorResult('Invalid regex for group_by: ' + groupBy);
+          }
+          extractGroup = function(line) {
+            var match = line.match(customRegex);
+            if (!match) return null;
+            return match[1] || match[0]; // Prefer capture group, fall back to full match
+          };
+        }
+
+        var counts: Record<string, number> = {};
+        var totalLines = 0;
+
+        for (var file of files) {
+          try {
+            await readLines(file, function(line, _lineNum, _filePath) {
+              totalLines++;
+              var timestamp = extractTimestamp(line);
+              if (!isWithinTimeRange(timestamp, since, until)) return;
+              var group = extractGroup(line);
+              if (group) {
+                counts[group] = (counts[group] || 0) + 1;
+              }
+            });
+          } catch { /* skip */ }
+        }
+
+        // Sort by count descending
+        var sorted = Object.entries(counts)
+          .sort(function(a, b) { return b[1] - a[1]; })
+          .map(function(entry) { return { group: entry[0], count: entry[1] }; });
+
+        if (sorted.length === 0) {
+          return textResult('No groups found for "' + groupBy + '" in ' + totalLines + ' lines.');
+        }
+
+        return jsonResult({
+          groupBy: groupBy,
+          filesSearched: files.length,
+          totalLines: totalLines,
+          groups: sorted,
+        });
+      },
+    },
+
+    {
+      name: 'ent_log_tail',
+      label: 'Tail Log File',
+      description: 'Read the last N lines of a log file. Efficient for checking recent log entries.',
+      category: 'search',
+      risk: 'low',
+      parameters: {
+        type: 'object',
+        properties: {
+          path: { type: 'string', description: 'Path to the log file.' },
+          lines: { type: 'number', description: 'Number of lines to read from the end (default 50).' },
+        },
+        required: ['path'],
+      },
+      execute: async function(_toolCallId, args) {
+        var params = args as Record<string, unknown>;
+        var targetPath = resolvePath(readStringParam(params, 'path', { required: true }), options);
+        var numLines = readNumberParam(params, 'lines', { integer: true }) ?? DEFAULT_TAIL_LINES;
+
+        try {
+          var stat = await fsPromises.stat(targetPath);
+          if (stat.isDirectory()) {
+            return errorResult(targetPath + ' is a directory. Provide a specific log file path.');
+          }
+        } catch {
+          return errorResult('File not found: ' + targetPath);
+        }
+
+        try {
+          var content = await fsPromises.readFile(targetPath, 'utf-8');
+          var allLines = content.split('\n');
+          // Remove trailing empty line if present
+          if (allLines.length > 0 && allLines[allLines.length - 1] === '') {
+            allLines.pop();
+          }
+          var startIdx = Math.max(0, allLines.length - numLines);
+          var tailLines = allLines.slice(startIdx);
+
+          var output = tailLines.map(function(line, i) {
+            var lineNum = startIdx + i + 1;
+            return lineNum + '\t' + line;
+          }).join('\n');
+
+          return textResult(
+            'Last ' + tailLines.length + ' lines of ' + targetPath +
+            ' (total: ' + allLines.length + ' lines)\n\n' + output
+          );
+        } catch (err: any) {
+          return errorResult('Failed to read file: ' + (err.message || targetPath));
+        }
+      },
+    },
+
+    {
+      name: 'ent_log_correlate',
+      label: 'Correlate Log Entries',
+      description: 'Find related log entries by correlation ID, request ID, or trace ID. Searches across all log files in a directory.',
+      category: 'search',
+      risk: 'low',
+      parameters: {
+        type: 'object',
+        properties: {
+          path: { type: 'string', description: 'Log file or directory path.' },
+          correlation_id: { type: 'string', description: 'Correlation ID, request ID, or trace ID to search for.' },
+          limit: { type: 'number', description: 'Maximum entries to return (default 50).' },
+        },
+        required: ['path', 'correlation_id'],
+      },
+      execute: async function(_toolCallId, args) {
+        var params = args as Record<string, unknown>;
+        var targetPath = resolvePath(readStringParam(params, 'path', { required: true }), options);
+        var correlationId = readStringParam(params, 'correlation_id', { required: true });
+        var limit = readNumberParam(params, 'limit', { integer: true }) ?? DEFAULT_CORRELATE_LIMIT;
+
+        var files = await collectLogFiles(targetPath);
+        if (files.length === 0) {
+          return errorResult('No log files found at: ' + targetPath);
+        }
+
+        // Escape special regex characters in the correlation ID
+        var escapedId = correlationId.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
+        var regex = new RegExp(escapedId, 'i');
+
+        var entries: Array<{ file: string; line: number; text: string; timestamp: string | null }> = [];
+
+        for (var file of files) {
+          if (entries.length >= limit) break;
+          try {
+            await readLines(file, function(line, lineNum, filePath) {
+              if (entries.length >= limit) return false;
+              if (regex.test(line)) {
+                entries.push({
+                  file: path.relative(resolvePath('.', options), filePath),
+                  line: lineNum,
+                  text: line.length > 500 ? line.slice(0, 500) + '...' : line,
+                  timestamp: extractTimestamp(line),
+                });
+              }
+            });
+          } catch { /* skip unreadable */ }
+        }
+
+        if (entries.length === 0) {
+          return textResult('No log entries found for correlation ID: ' + correlationId);
+        }
+
+        // Sort by timestamp if available
+        entries.sort(function(a, b) {
+          if (a.timestamp && b.timestamp) return a.timestamp.localeCompare(b.timestamp);
+          if (a.timestamp) return -1;
+          if (b.timestamp) return 1;
+          return 0;
+        });
+
+        return jsonResult({
+          correlationId: correlationId,
+          filesSearched: files.length,
+          entryCount: entries.length,
+          truncated: entries.length >= limit,
+          entries: entries,
+        });
+      },
+    },
+
+    {
+      name: 'ent_log_errors',
+      label: 'Extract Log Errors',
+      description: 'Extract error-level entries (ERROR, FATAL, CRITICAL, Exception, Traceback, panic) from logs with surrounding context lines.',
+      category: 'search',
+      risk: 'low',
+      parameters: {
+        type: 'object',
+        properties: {
+          path: { type: 'string', description: 'Log file or directory path.' },
+          since: { type: 'string', description: 'Start time filter (ISO date, optional).' },
+          limit: { type: 'number', description: 'Maximum error entries to return (default 50).' },
+        },
+        required: ['path'],
+      },
+      execute: async function(_toolCallId, args) {
+        var params = args as Record<string, unknown>;
+        var targetPath = resolvePath(readStringParam(params, 'path', { required: true }), options);
+        var since = readStringParam(params, 'since');
+        var limit = readNumberParam(params, 'limit', { integer: true }) ?? DEFAULT_ERROR_LIMIT;
+
+        var files = await collectLogFiles(targetPath);
+        if (files.length === 0) {
+          return errorResult('No log files found at: ' + targetPath);
+        }
+
+        var errors: Array<{
+          file: string;
+          line: number;
+          text: string;
+          context: string[];
+          timestamp: string | null;
+        }> = [];
+
+        for (var file of files) {
+          if (errors.length >= limit) break;
+          try {
+            // Read file into line buffer for context extraction
+            var content = await fsPromises.readFile(file, 'utf-8');
+            var allLines = content.split('\n');
+
+            for (var i = 0; i < allLines.length; i++) {
+              if (errors.length >= limit) break;
+              var line = allLines[i];
+              if (!ERROR_PATTERNS.test(line)) continue;
+
+              var timestamp = extractTimestamp(line);
+              if (since && timestamp && timestamp < since) continue;
+
+              // Gather context: 1 line before, 2 lines after
+              var contextLines: string[] = [];
+              if (i > 0) contextLines.push('[' + i + '] ' + allLines[i - 1]);
+              contextLines.push('[' + (i + 1) + '] ' + line);
+              if (i + 1 < allLines.length) contextLines.push('[' + (i + 2) + '] ' + allLines[i + 1]);
+              if (i + 2 < allLines.length) contextLines.push('[' + (i + 3) + '] ' + allLines[i + 2]);
+
+              errors.push({
+                file: path.relative(resolvePath('.', options), file),
+                line: i + 1,
+                text: line.length > 500 ? line.slice(0, 500) + '...' : line,
+                context: contextLines,
+                timestamp: timestamp,
+              });
+            }
+          } catch { /* skip unreadable */ }
+        }
+
+        if (errors.length === 0) {
+          var msg = 'No error entries found in ' + files.length + ' file(s)';
+          if (since) msg += ' since ' + since;
+          return textResult(msg + '.');
+        }
+
+        return jsonResult({
+          filesSearched: files.length,
+          errorCount: errors.length,
+          truncated: errors.length >= limit,
+          errors: errors,
+        });
+      },
+    },
+  ];
+}