linshenkx/prompt-optimizer · error · Error

Prompt must be a non-empty string

Error message

Prompt must be a non-empty string

What it means

ParameterAdapter.validatePrompt rejects a prompt argument that is falsy, not a string, or a string containing only whitespace. It runs in the MCP server request handlers, guarding the core prompt input before processing; a second check enforces the 50,000-character maximum.

Source

Thrown at packages/mcp-server/src/adapters/parameter-adapter.ts:13

/**
 * 参数验证工具
 * 简化的参数验证，移除过度抽象
 */

export class ParameterValidator {

  /**
   * 验证提示词输入
   */
  static validatePrompt(prompt: string): void {
    if (!prompt || typeof prompt !== 'string' || prompt.trim().length === 0) {
      throw new Error('Prompt must be a non-empty string');
    }
    if (prompt.length > 50000) {
      throw new Error('Prompt must not exceed 50,000 characters');
    }
  }

  /**
   * 验证模板输入
   */
  static validateTemplate(template?: string): void {
    if (template !== undefined && (typeof template !== 'string' || template.trim().length === 0)) {
      throw new Error('Template must be a non-empty string');
    }
  }

  /**
   * 验证需求描述输入
   */

View on GitHub (pinned to 3e677b1d9f)

Solutions

1. Ensure the caller always sends a non-empty trimmed string for prompt
2. Trim/validate client-side before invoking: if (!prompt?.trim()) fix the caller
3. Check the tool's input schema and the client's argument construction for typos in the parameter name
4. If undefined is legitimate for your flow, substitute a default prompt string before calling

Example fix

// before
await tool.call({ prompt: userInput }); // userInput may be ''

// after
await tool.call({ prompt: (userInput ?? '').trim() || fallbackPrompt });

if (typeof prompt !== 'string' || prompt.trim().length === 0) {
  throw new Error('client-side: prompt required');
}

const isValidPrompt = (p: unknown): p is string =>
  typeof p === 'string' && p.trim().length > 0;

try {
  await server.callTool('prompt', { prompt });
} catch (e) {
  if ((e as Error).message.includes('Prompt must be a non-empty string')) {
    return { error: 'Please provide a prompt' };
  }
  throw e;
}

When it happens

Trigger: Calling the MCP tool/prompt endpoint with prompt: '' , prompt: null, prompt: 123 (wrong type), or prompt: ' ' (whitespace only). These typically come from an LLM client constructing tool-call arguments incorrectly.

Common situations: An MCP client (Claude, Cursor, custom integration) omitting the prompt field or sending an empty string; templating bugs producing undefined interpolated into the call; schema drift where the client sends a different parameter name so prompt arrives undefined.

Related errors

• Prompt must not exceed 50,000 characters
• Template must be a non-empty string
• Requirements must be a non-empty string
• Iteration requirement must not be empty.
• ${label} promptText must not be empty.

AI-assisted analysis of linshenkx/prompt-optimizer@3e677b1d9f (2026-08-27). Data as JSON: /api/errors/16d1375906aa23a5. Report an issue: GitHub.