Skip to content
AgentForceZone

.prompt()

.prompt()

Section titled “.prompt()”

AgentForceAgent Required Chainable

Set the main user prompt that defines the task, question, or instruction for the agent to process. This is the primary input that the agent will respond to based on its system prompt and configuration.

Signature

Section titled “Signature”
prompt(prompt: string): AgentForceAgent

Parameters

Section titled “Parameters”
Parameter
Type
Default
Description
prompt
string
undefined
The user prompt text that defines the task or question for the agent

Returns

Section titled “Returns”

Returns the AgentForceAgent instance for method chaining.

Examples:

Section titled “Examples:”

Basic Usage

Section titled “Basic Usage”
import { AgentForceAgent } from '@agentforce/adk';


const agent = new AgentForceAgent({
  name: "QuestionAgent"
})
  .useLLM("ollama", "gemma3:4b")
  .prompt("What is the capital of France?");


const response = await agent.output("text");
console.log(response); // "Paris is the capital of France."

Complex Prompts

Section titled “Complex Prompts”
const agent = new AgentForceAgent({ name: "ComplexAgent" })
  .systemPrompt("You are a senior software architect")
  .prompt(`
    I need help designing a user authentication system with these requirements:


    1. Support for email/password and social login (Google, GitHub)
    2. JWT-based session management
    3. Role-based access control (admin, user, guest)
    4. Password reset functionality
    5. Account verification via email


    Please provide:
    - Database schema design
    - API endpoint structure
    - Security best practices
    - Implementation recommendations
  `);

Dynamic Prompts

Section titled “Dynamic Prompts”

Build prompts dynamically based on user input or data:

// Function to generate dynamic prompts
function createAnalysisPrompt(dataType: string, data: any[], analysisType: string) {
  return `
    Please perform ${analysisType} analysis on this ${dataType} data:


    Data: ${JSON.stringify(data, null, 2)}


    Focus on:
    - Key trends and patterns
    - Notable anomalies or outliers
    - Actionable insights
    - Recommendations for next steps
  `;
}


// Usage with different data types
const salesAgent = new AgentForceAgent({ name: "SalesAnalyst" })
  .systemPrompt("You are a business analyst expert")
  .prompt(createAnalysisPrompt("sales", salesData, "quarterly"));


const userAgent = new AgentForceAgent({ name: "UserAnalyst" })
  .systemPrompt("You are a UX research specialist")
  .prompt(createAnalysisPrompt("user behavior", userData, "engagement"));

Method Chaining

Section titled “Method Chaining”

The prompt() method supports fluent chaining and is typically used after system configuration:

const agent = new AgentForceAgent({ name: "ChainedAgent" })
  .useLLM("google", "gemini-1.5-pro")
  .systemPrompt("You are a technical documentation specialist")
  .prompt("Explain how async/await works in JavaScript with examples")
  .debug(); // Enable debug logging


const response = await agent.output("text");

Use Cases

Section titled “Use Cases”

Question Answering

Prompt Pattern: Direct questions or knowledge queries

Best for: FAQs, explanations, factual information

Example: “Explain the difference between REST and GraphQL APIs”

Code Generation

Prompt Pattern: Specific implementation requests with requirements

Best for: Functions, components, algorithms, boilerplate code

Example: “Create a React component for a user profile card”

Content Creation

Prompt Pattern: Creative briefs with style and format specifications

Best for: Blog posts, documentation, marketing copy

Example: “Write a technical blog post about Docker containers”

Data Analysis

Prompt Pattern: Data + analysis instructions + desired output format

Best for: Reports, insights, trend analysis, recommendations

Example: “Analyze these sales metrics and identify growth opportunities”

Best Practices

Section titled “Best Practices”

✅ Effective Prompts

Section titled “✅ Effective Prompts”
// Clear and specific
.prompt("Generate a TypeScript interface for a user object with name, email, age, and optional bio fields")


// Provide context and constraints
.prompt(`Create a responsive navbar component using:
- React with TypeScript
- Tailwind CSS for styling
- Support for mobile hamburger menu
- Include logo, navigation links, and user dropdown`)


// Specify desired output format
.prompt("List the top 5 JavaScript frameworks and explain each in exactly 2 sentences")


// Include examples when helpful
.prompt(`Convert this SQL query to MongoDB aggregation:
SELECT category, COUNT(*) as count, AVG(price) as avg_price
FROM products
WHERE price > 100
GROUP BY category
ORDER BY count DESC`)

❌ Common Issues

Section titled “❌ Common Issues”
// Too vague
.prompt("Help me with my code")


// No context or requirements
.prompt("Make this better")


// Conflicting instructions
.prompt("Write a short but comprehensive detailed guide")


// Missing critical information
.prompt("Fix this bug") // What bug? What code?

Error Handling

Section titled “Error Handling”
function createAgentWithPrompt(userInput: string) {
  // Validate input
  if (!userInput || typeof userInput !== 'string') {
    throw new Error('Valid prompt string is required');
  }


  if (userInput.trim().length === 0) {
    throw new Error('Prompt cannot be empty');
  }


  try {
    const agent = new AgentForceAgent({ name: "SafeAgent" })
      .systemPrompt("You are a helpful assistant")
      .prompt(userInput.trim());


    return agent;
  } catch (error) {
    console.error("Failed to create agent:", error.message);
    throw error;
  }
}


// Usage with error handling
try {
  const agent = createAgentWithPrompt(userInput);
  const response = await agent.output("text");
} catch (error) {
  console.error("Agent Error:", error.message);
}
Section titled “Related Methods”

See Also

Section titled “See Also”