Example Agent Configurations
Learn from complete, working agent configurations. Copy these examples and customize them for your own agents.
Simple Text Agent
A basic agent that provides information without requiring complex inputs.
Use Case: Daily affirmations, simple Q&A, fact retrieval
{
"metadata": {
"name": "Daily Affirmation Bot",
"description": "Start your day with a positive affirmation",
"category": "wellness",
"tags": ["affirmation", "motivation", "daily", "positivity"],
"version": "1.0.0",
"accessLevel": "public"
},
"execution": {
"model": "gpt-4o-mini",
"systemPrompt": "You are a wellness coach specializing in positive affirmations. Create unique, personalized affirmations that are uplifting and empowering. Each affirmation should be present-tense, positive, and specific.",
"temperature": 0.9,
"maxTokens": 150,
"responseFormat": "text"
},
"schemas": {
"inputSchema": {
"type": "object",
"properties": {
"focus": {
"type": "string",
"description": "What area would you like to focus on?",
"enum": ["confidence", "health", "relationships", "career", "creativity"],
"ui:widget": "select"
}
}
},
"outputSchema": {
"type": "object",
"properties": {
"affirmation": {
"type": "string"
}
}
}
}
}
Content Creator Agent
An agent that generates content with multiple output fields.
Use Case: Blog posts, social media, marketing copy
{
"metadata": {
"name": "Social Media Content Generator",
"description": "Create engaging social media posts with hashtags",
"category": "marketing",
"logo": "https://example.com/social-media-logo.png",
"tags": ["social media", "marketing", "content", "instagram", "twitter"],
"version": "1.2.0",
"accessLevel": "public"
},
"execution": {
"model": "gpt-4o",
"systemPrompt": "You are a social media expert who creates engaging, platform-optimized content. For each request:\n1. Write compelling copy that hooks readers\n2. Include relevant hashtags (5-10 for Instagram, 2-3 for Twitter)\n3. Suggest the best posting time\n4. Keep within character limits\n\nTone should match the platform and brand voice requested.",
"temperature": 0.8,
"maxTokens": 500,
"responseFormat": "json"
},
"schemas": {
"inputSchema": {
"type": "object",
"properties": {
"platform": {
"type": "string",
"description": "Which platform?",
"enum": ["Instagram", "Twitter", "LinkedIn", "Facebook", "TikTok"],
"ui:widget": "select"
},
"topic": {
"type": "string",
"description": "What's your post about?",
"ui:widget": "textarea",
"ui:placeholder": "e.g., launching a new product, sharing a tip, behind-the-scenes"
},
"tone": {
"type": "string",
"description": "What tone should the post have?",
"enum": ["professional", "casual", "funny", "inspirational", "educational"],
"ui:widget": "select"
},
"callToAction": {
"type": "boolean",
"description": "Include a call-to-action?",
"default": true
}
},
"required": ["platform", "topic", "tone"]
},
"outputSchema": {
"type": "object",
"properties": {
"caption": {
"type": "string",
"description": "The main post text"
},
"hashtags": {
"type": "array",
"items": { "type": "string" },
"description": "Relevant hashtags"
},
"bestTime": {
"type": "string",
"description": "Suggested posting time"
},
"tips": {
"type": "string",
"description": "Platform-specific tips"
}
}
}
}
}
Data Analysis Agent
An agent that processes structured input and provides insights.
Use Case: Report generation, data interpretation, analytics
{
"metadata": {
"name": "Fitness Progress Analyzer",
"description": "Analyze your workout data and get insights",
"category": "health_fitness",
"tags": ["fitness", "analytics", "health", "workout", "progress"],
"version": "1.0.0"
},
"execution": {
"model": "gpt-4o",
"systemPrompt": "You are a fitness analyst who reviews workout data and provides actionable insights. Analyze the user's progress, identify patterns, celebrate achievements, and provide specific recommendations for improvement. Always be encouraging and data-driven.",
"temperature": 0.3,
"maxTokens": 800,
"responseFormat": "markdown"
},
"schemas": {
"inputSchema": {
"type": "object",
"properties": {
"workoutType": {
"type": "string",
"description": "What type of workout?",
"enum": ["strength", "cardio", "yoga", "sports", "mixed"]
},
"duration": {
"type": "number",
"description": "Duration in minutes"
},
"frequency": {
"type": "number",
"description": "Workouts per week"
},
"goal": {
"type": "string",
"description": "Your fitness goal",
"enum": ["lose weight", "build muscle", "improve endurance", "increase flexibility", "general fitness"]
},
"challenges": {
"type": "string",
"description": "Any challenges you're facing?",
"ui:widget": "textarea",
"ui:placeholder": "e.g., lack of motivation, time constraints, plateauing"
}
},
"required": ["workoutType", "duration", "frequency", "goal"]
},
"outputSchema": {
"type": "object",
"properties": {
"summary": {
"type": "string",
"description": "Overall assessment"
},
"achievements": {
"type": "array",
"items": { "type": "string" },
"description": "What you're doing well"
},
"recommendations": {
"type": "array",
"items": { "type": "string" },
"description": "Specific actions to improve"
},
"nextSteps": {
"type": "string",
"description": "Your action plan"
}
}
}
}
}
Research Assistant Agent
An agent with web search capabilities for research tasks.
Use Case: Information gathering, fact-checking, research
{
"metadata": {
"name": "Research Assistant Pro",
"description": "Deep dive into any topic with web search and analysis",
"category": "research",
"tags": ["research", "analysis", "web search", "information", "facts"],
"version": "2.0.0"
},
"execution": {
"model": "gpt-4o",
"systemPrompt": "You are an expert research assistant. When given a topic:\n1. Use web search to find current, reliable information\n2. Synthesize findings from multiple sources\n3. Cite sources clearly\n4. Present information in an organized, easy-to-understand format\n5. Distinguish between facts and opinions\n6. Highlight any conflicting information found",
"temperature": 0.4,
"maxTokens": 2000,
"responseFormat": "markdown",
"tools": [
{
"type": "web_search",
"enabled": true,
"maxResults": 10
}
]
},
"schemas": {
"inputSchema": {
"type": "object",
"properties": {
"topic": {
"type": "string",
"description": "What do you want to research?",
"ui:widget": "textarea",
"ui:placeholder": "e.g., latest developments in renewable energy"
},
"depth": {
"type": "string",
"description": "How detailed should the research be?",
"enum": ["quick overview", "moderate detail", "comprehensive deep dive"],
"default": "moderate detail"
},
"focusArea": {
"type": "string",
"description": "Any specific aspect to focus on?",
"ui:placeholder": "e.g., environmental impact, cost analysis, future trends"
}
},
"required": ["topic"]
},
"outputSchema": {
"type": "object",
"properties": {
"summary": {
"type": "string",
"description": "Executive summary"
},
"keyFindings": {
"type": "array",
"items": { "type": "string" },
"description": "Main discoveries"
},
"detailedAnalysis": {
"type": "string",
"description": "In-depth information"
},
"sources": {
"type": "array",
"items": { "type": "string" },
"description": "Referenced sources"
}
}
}
}
}
Creative Writing Agent
An agent optimized for storytelling and creative content.
Use Case: Story writing, character development, creative prompts
{
"metadata": {
"name": "Story Weaver",
"description": "Your creative writing companion for stories and narratives",
"category": "writing",
"tags": ["creative writing", "stories", "fiction", "narrative", "characters"],
"version": "1.1.0"
},
"execution": {
"model": "claude-3-5-sonnet",
"systemPrompt": "You are a creative writing mentor and storyteller. Help users develop compelling narratives with:\n- Vivid descriptions and imagery\n- Well-developed characters\n- Engaging dialogue\n- Proper story structure (beginning, middle, end)\n- Literary devices and techniques\n\nAdapt your style to match the user's chosen genre and tone.",
"temperature": 1.2,
"maxTokens": 1500,
"responseFormat": "markdown"
},
"schemas": {
"inputSchema": {
"type": "object",
"properties": {
"genre": {
"type": "string",
"description": "Story genre",
"enum": ["fantasy", "sci-fi", "mystery", "romance", "horror", "thriller", "literary fiction", "adventure"],
"ui:widget": "select"
},
"prompt": {
"type": "string",
"description": "Story idea or prompt",
"ui:widget": "textarea",
"ui:placeholder": "e.g., A detective discovers they can hear thoughts, but only lies"
},
"length": {
"type": "string",
"description": "Desired story length",
"enum": ["flash fiction (500 words)", "short story (1000 words)", "chapter (2000 words)"],
"default": "short story (1000 words)"
},
"tone": {
"type": "string",
"description": "Story tone",
"enum": ["dark", "lighthearted", "serious", "humorous", "melancholic", "hopeful"]
}
},
"required": ["genre", "prompt"]
},
"outputSchema": {
"type": "object",
"properties": {
"story": {
"type": "string",
"description": "The complete story"
},
"writingNotes": {
"type": "string",
"description": "Literary techniques used"
},
"suggestions": {
"type": "string",
"description": "Ideas for expanding the story"
}
}
}
}
}
Meal Planning Agent
An agent that provides structured, practical advice.
Use Case: Meal planning, recipes, nutrition
{
"metadata": {
"name": "Smart Meal Planner",
"description": "Personalized meal plans based on your preferences and goals",
"category": "health_fitness",
"tags": ["nutrition", "meals", "recipes", "diet", "healthy eating"],
"version": "1.3.0"
},
"execution": {
"model": "gpt-4o-mini",
"systemPrompt": "You are a nutritionist and meal planning expert. Create practical, delicious meal plans that:\n- Match dietary restrictions and preferences\n- Meet nutritional goals\n- Use accessible, affordable ingredients\n- Include preparation tips\n- Provide variety and balance\n\nAlways consider portion sizes and macronutrients.",
"temperature": 0.7,
"maxTokens": 1200,
"responseFormat": "markdown"
},
"schemas": {
"inputSchema": {
"type": "object",
"properties": {
"dietType": {
"type": "string",
"description": "Dietary preference",
"enum": ["omnivore", "vegetarian", "vegan", "pescatarian", "keto", "paleo", "Mediterranean"],
"ui:widget": "select"
},
"meals": {
"type": "array",
"description": "Which meals to plan?",
"items": {
"type": "string",
"enum": ["breakfast", "lunch", "dinner", "snacks"]
},
"default": ["breakfast", "lunch", "dinner"]
},
"allergies": {
"type": "string",
"description": "Any allergies or foods to avoid?",
"ui:placeholder": "e.g., nuts, dairy, shellfish"
},
"goal": {
"type": "string",
"description": "Nutrition goal",
"enum": ["weight loss", "muscle gain", "maintenance", "energy boost", "general health"]
},
"cookingTime": {
"type": "string",
"description": "Available cooking time",
"enum": ["under 15 min", "15-30 min", "30-60 min", "no limit"]
}
},
"required": ["dietType", "goal"]
},
"outputSchema": {
"type": "object",
"properties": {
"mealPlan": {
"type": "string",
"description": "Complete meal plan with recipes"
},
"shoppingList": {
"type": "array",
"items": { "type": "string" },
"description": "Ingredients needed"
},
"nutritionSummary": {
"type": "string",
"description": "Macro breakdown and calories"
},
"prepTips": {
"type": "string",
"description": "Meal prep suggestions"
}
}
}
}
}
Code Helper Agent
An agent for programming assistance.
Use Case: Code generation, debugging, explanation
{
"metadata": {
"name": "Code Mentor",
"description": "Your programming assistant for coding help and debugging",
"category": "code",
"tags": ["programming", "coding", "debugging", "development", "software"],
"version": "2.1.0"
},
"execution": {
"model": "gpt-4o",
"systemPrompt": "You are an expert programmer and teacher. Help users with:\n- Writing clean, efficient code\n- Debugging issues\n- Explaining concepts clearly\n- Best practices and patterns\n- Code optimization\n\nAlways:\n- Provide working code examples\n- Explain your reasoning\n- Suggest improvements\n- Follow language conventions",
"temperature": 0.2,
"maxTokens": 2000,
"responseFormat": "markdown"
},
"schemas": {
"inputSchema": {
"type": "object",
"properties": {
"language": {
"type": "string",
"description": "Programming language",
"enum": ["JavaScript", "Python", "Java", "C++", "Go", "Rust", "TypeScript", "Swift", "PHP"],
"ui:widget": "select"
},
"taskType": {
"type": "string",
"description": "What do you need help with?",
"enum": ["write new code", "debug existing code", "optimize code", "explain concept", "code review"]
},
"description": {
"type": "string",
"description": "Describe your coding task or problem",
"ui:widget": "textarea",
"ui:placeholder": "e.g., I need a function to validate email addresses"
},
"code": {
"type": "string",
"description": "Existing code (if any)",
"ui:widget": "textarea",
"ui:placeholder": "Paste your code here..."
}
},
"required": ["language", "taskType", "description"]
},
"outputSchema": {
"type": "object",
"properties": {
"solution": {
"type": "string",
"description": "Code solution with explanation"
},
"bestPractices": {
"type": "string",
"description": "Recommendations and tips"
},
"resources": {
"type": "array",
"items": { "type": "string" },
"description": "Additional learning resources"
}
}
}
}
}
Tips for Customizing Examples
1. Adjust the System Prompt
The system prompt is the most important customization. Make it specific to your use case.
2. Modify Input Fields
Add, remove, or change input fields to match what your agent needs.
3. Choose the Right Model
- Use GPT-4o for complex reasoning
- Use GPT-4o-mini for quick, simple tasks
- Use Claude for creative writing
4. Set Appropriate Temperature
- Low (0-0.3): Factual, consistent responses
- Medium (0.4-0.8): Balanced creativity
- High (0.9-2.0): Very creative, varied outputs
5. Add UI Hints
Use ui:widget, ui:placeholder, and other hints to improve user experience.
Next Steps
- Review Configuration Guide for detailed field explanations
- Check JSON Schema Reference for schema options
- Create Your Agent using these examples
- Share your agent in the Agents Marketplace
Need More Examples?
Browse the Agents Marketplace to see hundreds of live agents and their configurations.