diff --git a/src/utils/errors.js b/src/utils/errors.js
new file mode 100644
index 00000000..80473917
--- /dev/null
+++ b/src/utils/errors.js
@@ -0,0 +1,204 @@
+/**
+ * Error Classification and User-Friendly Messages
+ *
+ * Provides utilities for classifying errors and generating
+ * helpful error messages for users.
+ */
+
+/**
+ * Error type classifications
+ */
+export const ErrorType = {
+  // Network-related errors
+  NETWORK: 'network',
+  TIMEOUT: 'timeout',
+
+  // API errors
+  API_ERROR: 'api_error',
+  API_RATE_LIMIT: 'api_rate_limit',
+  API_UNAUTHORIZED: 'api_unauthorized',
+  API_NOT_FOUND: 'api_not_found',
+  API_SERVER_ERROR: 'api_server_error',
+
+  // Discord-specific errors
+  DISCORD_PERMISSION: 'discord_permission',
+  DISCORD_CHANNEL_NOT_FOUND: 'discord_channel_not_found',
+  DISCORD_MISSING_ACCESS: 'discord_missing_access',
+
+  // Configuration errors
+  CONFIG_MISSING: 'config_missing',
+  CONFIG_INVALID: 'config_invalid',
+
+  // Unknown/generic errors
+  UNKNOWN: 'unknown',
+};
+
+/**
+ * Classify an error into a specific error type
+ *
+ * @param {Error} error - The error to classify
+ * @param {Object} context - Optional context (response, statusCode, etc.)
+ * @returns {string} Error type from ErrorType enum
+ */
+export function classifyError(error, context = {}) {
+  if (!error) return ErrorType.UNKNOWN;
+
+  const message = error.message?.toLowerCase() || '';
+  const code = error.code || context.code;
+  const status = error.status || context.status || context.statusCode;
+
+  // Network errors
+  if (code === 'ECONNREFUSED' || code === 'ENOTFOUND' || code === 'ETIMEDOUT') {
+    return ErrorType.NETWORK;
+  }
+  if (code === 'ETIMEDOUT' || message.includes('timeout')) {
+    return ErrorType.TIMEOUT;
+  }
+  if (message.includes('fetch failed') || message.includes('network')) {
+    return ErrorType.NETWORK;
+  }
+
+  // HTTP status code errors
+  if (status) {
+    if (status === 401 || status === 403) {
+      return ErrorType.API_UNAUTHORIZED;
+    }
+    if (status === 404) {
+      return ErrorType.API_NOT_FOUND;
+    }
+    if (status === 429) {
+      return ErrorType.API_RATE_LIMIT;
+    }
+    if (status >= 500) {
+      return ErrorType.API_SERVER_ERROR;
+    }
+    if (status >= 400) {
+      return ErrorType.API_ERROR;
+    }
+  }
+
+  // Discord-specific errors
+  if (code === 50001 || message.includes('missing access')) {
+    return ErrorType.DISCORD_MISSING_ACCESS;
+  }
+  if (code === 50013 || message.includes('missing permissions')) {
+    return ErrorType.DISCORD_PERMISSION;
+  }
+  if (code === 10003 || message.includes('unknown channel')) {
+    return ErrorType.DISCORD_CHANNEL_NOT_FOUND;
+  }
+
+  // Config errors
+  if (message.includes('config.json not found') || message.includes('enoent')) {
+    return ErrorType.CONFIG_MISSING;
+  }
+  if (message.includes('invalid') && message.includes('config')) {
+    return ErrorType.CONFIG_INVALID;
+  }
+
+  // API errors (generic)
+  if (message.includes('api error') || context.isApiError) {
+    return ErrorType.API_ERROR;
+  }
+
+  return ErrorType.UNKNOWN;
+}
+
+/**
+ * Get a user-friendly error message based on error type
+ *
+ * @param {Error} error - The error object
+ * @param {Object} context - Optional context for more specific messages
+ * @returns {string} User-friendly error message
+ */
+export function getUserFriendlyMessage(error, context = {}) {
+  const errorType = classifyError(error, context);
+
+  const messages = {
+    [ErrorType.NETWORK]: "I'm having trouble connecting to my brain right now. Check if the AI service is running and try again!",
+
+    [ErrorType.TIMEOUT]: "That took too long to process. Try again with a shorter message, or wait a moment and retry!",
+
+    [ErrorType.API_RATE_LIMIT]: "Whoa, too many requests! Let's take a quick breather. Try again in a minute.",
+
+    [ErrorType.API_UNAUTHORIZED]: "I'm having authentication issues with the AI service. An admin needs to check the API credentials.",
+
+    [ErrorType.API_NOT_FOUND]: "The AI service endpoint isn't responding. Please check if it's configured correctly.",
+
+    [ErrorType.API_SERVER_ERROR]: "The AI service is having technical difficulties. It should recover automatically - try again in a moment!",
+
+    [ErrorType.API_ERROR]: "Something went wrong with the AI service. Give it another shot in a moment!",
+
+    [ErrorType.DISCORD_PERMISSION]: "I don't have permission to do that! An admin needs to check my role permissions.",
+
+    [ErrorType.DISCORD_CHANNEL_NOT_FOUND]: "I can't find that channel. It might have been deleted, or I don't have access to it.",
+
+    [ErrorType.DISCORD_MISSING_ACCESS]: "I don't have access to that resource. Please check my permissions!",
+
+    [ErrorType.CONFIG_MISSING]: "Configuration file not found! Please create a config.json file (you can copy from config.example.json).",
+
+    [ErrorType.CONFIG_INVALID]: "The configuration file has errors. Please check config.json for syntax errors or missing required fields.",
+
+    [ErrorType.UNKNOWN]: "Something unexpected happened. Try again, and if it keeps happening, check the logs for details.",
+  };
+
+  return messages[errorType] || messages[ErrorType.UNKNOWN];
+}
+
+/**
+ * Get suggested next steps for an error
+ *
+ * @param {Error} error - The error object
+ * @param {Object} context - Optional context
+ * @returns {string|null} Suggested next steps or null if none
+ */
+export function getSuggestedNextSteps(error, context = {}) {
+  const errorType = classifyError(error, context);
+
+  const suggestions = {
+    [ErrorType.NETWORK]: "Make sure the AI service (OpenClaw) is running and accessible.",
+
+    [ErrorType.TIMEOUT]: "Try a shorter message or wait a moment before retrying.",
+
+    [ErrorType.API_RATE_LIMIT]: "Wait 60 seconds before trying again.",
+
+    [ErrorType.API_UNAUTHORIZED]: "Check the OPENCLAW_TOKEN environment variable and API credentials.",
+
+    [ErrorType.API_NOT_FOUND]: "Verify the OPENCLAW_URL environment variable points to the correct endpoint.",
+
+    [ErrorType.API_SERVER_ERROR]: "The service should recover automatically. If it persists, restart the AI service.",
+
+    [ErrorType.DISCORD_PERMISSION]: "Grant the bot appropriate permissions in Server Settings > Roles.",
+
+    [ErrorType.DISCORD_CHANNEL_NOT_FOUND]: "Update the channel ID in config.json or verify the channel exists.",
+
+    [ErrorType.DISCORD_MISSING_ACCESS]: "Ensure the bot has access to the required channels and roles.",
+
+    [ErrorType.CONFIG_MISSING]: "Create config.json from config.example.json and fill in your settings.",
+
+    [ErrorType.CONFIG_INVALID]: "Validate your config.json syntax using a JSON validator.",
+  };
+
+  return suggestions[errorType] || null;
+}
+
+/**
+ * Check if an error is retryable (transient failure)
+ *
+ * @param {Error} error - The error to check
+ * @param {Object} context - Optional context
+ * @returns {boolean} True if the error should be retried
+ */
+export function isRetryable(error, context = {}) {
+  const errorType = classifyError(error, context);
+
+  // Only retry transient failures, not user/config errors
+  const retryableTypes = [
+    ErrorType.NETWORK,
+    ErrorType.TIMEOUT,
+    ErrorType.API_SERVER_ERROR,
+    ErrorType.API_RATE_LIMIT,
+  ];
+
+  return retryableTypes.includes(errorType);
+}
diff --git a/src/utils/retry.js b/src/utils/retry.js
new file mode 100644
index 00000000..1d14e187
--- /dev/null
+++ b/src/utils/retry.js
@@ -0,0 +1,130 @@
+/**
+ * Retry Utility with Exponential Backoff
+ *
+ * Provides utilities for retrying operations with configurable
+ * exponential backoff and integration with error classification.
+ */
+
+import { isRetryable, classifyError } from './errors.js';
+import { warn, error, debug } from '../logger.js';
+
+/**
+ * Sleep for a specified duration
+ * @param {number} ms - Milliseconds to sleep
+ * @returns {Promise<void>}
+ */
+function sleep(ms) {
+  return new Promise(resolve => setTimeout(resolve, ms));
+}
+
+/**
+ * Calculate delay with exponential backoff
+ * @param {number} attempt - Current attempt number (0-indexed)
+ * @param {number} baseDelay - Base delay in milliseconds
+ * @param {number} maxDelay - Maximum delay in milliseconds
+ * @returns {number} Delay in milliseconds
+ */
+function calculateBackoff(attempt, baseDelay, maxDelay) {
+  // Exponential backoff: baseDelay * 2^attempt
+  const delay = baseDelay * Math.pow(2, attempt);
+
+  // Cap at maxDelay
+  return Math.min(delay, maxDelay);
+}
+
+/**
+ * Retry an async operation with exponential backoff
+ *
+ * @param {Function} fn - Async function to retry
+ * @param {Object} options - Retry configuration options
+ * @param {number} options.maxRetries - Maximum number of retry attempts (default: 3)
+ * @param {number} options.baseDelay - Initial delay in milliseconds (default: 1000)
+ * @param {number} options.maxDelay - Maximum delay in milliseconds (default: 30000)
+ * @param {Function} options.shouldRetry - Custom function to determine if error is retryable
+ * @param {Object} options.context - Optional context for logging
+ * @returns {Promise<any>} Result of the function
+ * @throws {Error} Throws the last error if all retries fail
+ */
+export async function withRetry(fn, options = {}) {
+  const {
+    maxRetries = 3,
+    baseDelay = 1000,
+    maxDelay = 30000,
+    shouldRetry = isRetryable,
+    context = {},
+  } = options;
+
+  let lastError;
+
+  for (let attempt = 0; attempt <= maxRetries; attempt++) {
+    try {
+      // Execute the function
+      return await fn();
+    } catch (err) {
+      lastError = err;
+
+      // Check if we should retry
+      const errorType = classifyError(err, context);
+      const canRetry = shouldRetry(err, context);
+
+      // Log the error
+      if (attempt === 0) {
+        warn(`Operation failed: ${err.message}`, {
+          ...context,
+          errorType,
+          attempt: attempt + 1,
+          maxRetries: maxRetries + 1,
+        });
+      }
+
+      // If this was the last attempt or error is not retryable, throw
+      if (attempt >= maxRetries || !canRetry) {
+        if (!canRetry) {
+          error('Operation failed with non-retryable error', {
+            ...context,
+            errorType,
+            attempt: attempt + 1,
+            error: err.message,
+          });
+        } else {
+          error('Operation failed after all retries', {
+            ...context,
+            errorType,
+            totalAttempts: attempt + 1,
+            error: err.message,
+          });
+        }
+        throw err;
+      }
+
+      // Calculate backoff delay
+      const delay = calculateBackoff(attempt, baseDelay, maxDelay);
+
+      debug(`Retrying in ${delay}ms`, {
+        ...context,
+        attempt: attempt + 1,
+        maxRetries: maxRetries + 1,
+        delay,
+        errorType,
+      });
+
+      // Wait before retrying
+      await sleep(delay);
+    }
+  }
+
+  // Should never reach here, but just in case
+  throw lastError;
+}
+
+/**
+ * Create a retry wrapper with pre-configured options
+ *
+ * @param {Object} defaultOptions - Default retry options
+ * @returns {Function} Configured retry function
+ */
+export function createRetryWrapper(defaultOptions = {}) {
+  return (fn, options = {}) => {
+    return withRetry(fn, { ...defaultOptions, ...options });
+  };
+}