Skip to content

Enhanced ChatGPT Clone: Features Agents, MCP, Skills, DeepSeek, Anthropic, AWS, OpenAI, Responses API, Azure, Groq, o1, GPT-5, Mistral, OpenRouter, Vertex AI, Gemini, Artifacts, AI model switching, message search, Code Interpreter, langchain, DALL-E-3, OpenAPI Actions, Functions, Secure Multi-User Auth, Presets, open-source for self-hosting. Active

🚏 feat: Rate Limit Error handling in MCPConnection (#10921)

Danny Avilaef96ce2b4ba5
diff --git a/packages/api/src/mcp/__tests__/MCPConnection.test.ts b/packages/api/src/mcp/__tests__/MCPConnection.test.ts
new file mode 100644
index 0000000000..a034f889e9
--- /dev/null
+++ b/packages/api/src/mcp/__tests__/MCPConnection.test.ts
@@ -0,0 +1,178 @@
+/**
+ * Tests for MCPConnection error detection methods.
+ *
+ * These tests use standalone implementations that mirror the private methods in MCPConnection.
+ * This approach was chosen because MCPConnection requires complex dependencies (Client, transport)
+ * that are difficult to mock properly. The standalone implementations are kept in sync with
+ * the actual implementation in connection.ts.
+ *
+ * Alternative approaches considered:
+ * 1. Reflection/type casting - fragile and breaks with refactoring
+ * 2. Protected methods with test subclass - changes public API for testing
+ * 3. Integration tests - tested separately in the full MCP test suite
+ */
+describe('MCPConnection Error Detection', () => {
+ /**
+ * Standalone implementation of isRateLimitError for testing.
+ * This mirrors the private method in MCPConnection (connection.ts).
+ * Keep in sync with the actual implementation.
+ */
+ function isRateLimitError(error: unknown): boolean {
+ if (!error || typeof error !== 'object') {
+ return false;
+ }
+
+ // Check for error code
+ if ('code' in error) {
+ const code = (error as { code?: number }).code;
+ if (code === 429) {
+ return true;
+ }
+ }
+
+ // Check message for rate limit indicators
+ if ('message' in error && typeof error.message === 'string') {
+ const message = error.message.toLowerCase();
+ if (
+ message.includes('429') ||
+ message.includes('rate limit') ||
+ message.includes('too many requests')
+ ) {
+ return true;
+ }
+ }
+
+ return false;
+ }
+
+ /**
+ * Standalone implementation of isOAuthError for testing.
+ * This mirrors the private method in MCPConnection (connection.ts).
+ * Keep in sync with the actual implementation.
+ */
+ function isOAuthError(error: unknown): boolean {
+ if (!error || typeof error !== 'object') {
+ return false;
+ }
+
+ // Check for error code
+ if ('code' in error) {
+ const code = (error as { code?: number }).code;
+ if (code === 401 || code === 403) {
+ return true;
+ }
+ }
+
+ // Check message for various auth error indicators
+ if ('message' in error && typeof error.message === 'string') {
+ const message = error.message.toLowerCase();
+ // Check for 401 status
+ if (message.includes('401') || message.includes('non-200 status code (401)')) {
+ return true;
+ }
+ // Check for invalid_token (OAuth servers return this for expired/revoked tokens)
+ if (message.includes('invalid_token')) {
+ return true;
+ }
+ // Check for authentication required
+ if (message.includes('authentication required') || message.includes('unauthorized')) {
+ return true;
+ }
+ }
+
+ return false;
+ }
+
+ describe('isRateLimitError', () => {
+ it('should detect rate limit error by code 429', () => {
+ const error = { code: 429, message: 'Too many requests' };
+ expect(isRateLimitError(error)).toBe(true);
+ });
+
+ it('should detect rate limit error by message containing 429', () => {
+ const error = { message: 'Error POSTing to endpoint (HTTP 429): Too many requests' };
+ expect(isRateLimitError(error)).toBe(true);
+ });
+
+ it('should detect rate limit error by message containing "rate limit"', () => {
+ const error = { message: 'Rate limit exceeded, please try again later' };
+ expect(isRateLimitError(error)).toBe(true);
+ });
+
+ it('should detect rate limit error by message containing "too many requests"', () => {
+ const error = { message: 'Too many requests - slow down!' };
+ expect(isRateLimitError(error)).toBe(true);
+ });
+
+ it('should not detect rate limit for 401 errors', () => {
+ const error = { code: 401, message: 'Unauthorized' };
+ expect(isRateLimitError(error)).toBe(false);
+ });
+
+ it('should not detect rate limit for 500 errors', () => {
+ const error = { code: 500, message: 'Internal server error' };
+ expect(isRateLimitError(error)).toBe(false);
+ });
+
+ it('should not detect rate limit for null/undefined', () => {
+ expect(isRateLimitError(null)).toBe(false);
+ expect(isRateLimitError(undefined)).toBe(false);
+ });
+
+ it('should not detect rate limit for non-object errors', () => {
+ expect(isRateLimitError('string error')).toBe(false);
+ expect(isRateLimitError(123)).toBe(false);
+ });
+
+ it('should handle real-world StackOverflow rate limit error', () => {
+ const error = {
+ code: 429,
+ message:
+ 'Streamable HTTP error: Error POSTing to endpoint: <!DOCTYPE html><html>Too Many Requests</html>',
+ };
+ expect(isRateLimitError(error)).toBe(true);
+ });
+ });
+
+ describe('isOAuthError', () => {
+ it('should detect OAuth error by code 401', () => {
+ const error = { code: 401, message: 'Unauthorized' };
+ expect(isOAuthError(error)).toBe(true);
+ });
+
+ it('should detect OAuth error by code 403', () => {
+ const error = { code: 403, message: 'Forbidden' };
+ expect(isOAuthError(error)).toBe(true);
+ });
+
+ it('should detect OAuth error by message containing 401', () => {
+ const error = { message: 'Error POSTing to endpoint (HTTP 401): Unauthorized' };
+ expect(isOAuthError(error)).toBe(true);
+ });
+
+ it('should not detect OAuth error for 429 rate limit', () => {
+ const error = { code: 429, message: 'Too many requests' };
+ expect(isOAuthError(error)).toBe(false);
+ });
+
+ it('should detect OAuth error for invalid_token', () => {
+ const error = { message: 'The access token is invalid_token or expired' };
+ expect(isOAuthError(error)).toBe(true);
+ });
+ });
+
+ describe('error type differentiation', () => {
+ it('should correctly differentiate between rate limit and OAuth errors', () => {
+ const rateLimitError = { code: 429, message: 'Too many requests' };
+ const oauthError = { code: 401, message: 'Unauthorized' };
+
+ // Rate limit error should be detected as rate limit, not OAuth
+ expect(isRateLimitError(rateLimitError)).toBe(true);
+ expect(isOAuthError(rateLimitError)).toBe(false);
+
+ // OAuth error should be detected as OAuth, not rate limit
+ expect(isOAuthError(oauthError)).toBe(true);
+ expect(isRateLimitError(oauthError)).toBe(false);
+ });
+ });
+});
diff --git a/packages/api/src/mcp/connection.ts b/packages/api/src/mcp/connection.ts
index e97bf05c6c..8a458c098f 100644
--- a/packages/api/src/mcp/connection.ts
+++ b/packages/api/src/mcp/connection.ts
@@ -415,6 +415,24 @@ export class MCPConnection extends EventEmitter {
} catch (error) {
logger.error(`${this.getLogPrefix()} Reconnection attempt failed:`, error);
+ // Stop immediately if rate limited - retrying will only make it worse
+ if (this.isRateLimitError(error)) {
+ /**
+ * Rate limiting sets shouldStopReconnecting to prevent hammering the server.
+ * Silent return here (vs throw in connectClient) because we're already in
+ * error recovery mode - throwing would just add noise. The connection
+ * must be recreated to retry after rate limit lifts.
+ */
+ logger.warn(
+ `${this.getLogPrefix()} Rate limited (429), stopping reconnection attempts`,
+ );
+ logger.debug(
+ `${this.getLogPrefix()} Rate limit block is permanent for this connection instance`,
+ );
+ this.shouldStopReconnecting = true;
+ return;
+ }
+
if (
this.reconnectAttempts === this.MAX_RECONNECT_ATTEMPTS ||
(this.shouldStopReconnecting as boolean)
@@ -475,6 +493,25 @@ export class MCPConnection extends EventEmitter {
this.emit('connectionChange', 'connected');
this.reconnectAttempts = 0;
} catch (error) {
+ // Check if it's a rate limit error - stop immediately to avoid making it worse
+ if (this.isRateLimitError(error)) {
+ /**
+ * Rate limiting sets shouldStopReconnecting to prevent hammering the server.
+ * This is a permanent block for this connection instance - the connection
+ * must be recreated (e.g., by user re-initiating) to retry after rate limit lifts.
+ *
+ * We throw here (unlike handleReconnection which returns silently) because:
+ * - connectClient() is a public API - callers expect async errors to throw
+ * - Other errors in this catch block also throw for consistency
+ * - handleReconnection is private/internal error recovery, different context
+ */
+ logger.warn(`${this.getLogPrefix()} Rate limited (429), stopping connection attempts`);
+ this.shouldStopReconnecting = true;
+ this.connectionState = 'error';
+ this.emit('connectionChange', 'error');
+ throw error;
+ }
+
// Check if it's an OAuth authentication error
if (this.isOAuthError(error)) {
logger.warn(`${this.getLogPrefix()} OAuth authentication required`);
@@ -786,4 +823,36 @@ export class MCPConnection extends EventEmitter {
return false;
}
+
+ /**
+ * Checks if an error indicates rate limiting (HTTP 429).
+ * Rate limited requests should stop reconnection attempts to avoid making the situation worse.
+ */
+ private isRateLimitError(error: unknown): boolean {
+ if (!error || typeof error !== 'object') {
+ return false;
+ }
+
+ // Check for error code
+ if ('code' in error) {
+ const code = (error as { code?: number }).code;
+ if (code === 429) {
+ return true;
+ }
+ }
+
+ // Check message for rate limit indicators
+ if ('message' in error && typeof error.message === 'string') {
+ const message = error.message.toLowerCase();
+ if (
+ message.includes('429') ||
+ message.includes('rate limit') ||
+ message.includes('too many requests')
+ ) {
+ return true;
+ }
+ }
+
+ return false;
+ }
}