Workflow Templates

Get started quickly with AgentMap using our curated collection of ready-to-use workflow templates. Each template is designed to showcase specific capabilities and can be customized for your unique use cases.

AgentMap Template Library

Ready-to-use workflow templates to get you started quickly

Category:

Difficulty:

🤖 Automation

🟢 Beginner

Weather Notification Bot

Daily weather alerts with intelligent notifications based on conditions

Use Case: Get daily weather updates with contextual advice (umbrella reminders, outfit suggestions)

notificationsweatherdaily-automation

Required Agents:

llmecho

Example Output:

🌤️ NYC Weather Update: 72°F, partly cloudy. Perfect day for a walk! Light jacket recommended for evening.

Configuration Notes:

Replace 'input' agent with weather API integration for automation
Add scheduling for daily notifications
Customize prompt for regional preferences

View CSV Content

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
WeatherBot,GetWeather,,Get current weather data,input,AnalyzeWeather,,location,weather_data,Enter your location (e.g. New York City):
WeatherBot,AnalyzeWeather,,"{'temperature': 0.7, 'model': 'gpt-3.5-turbo'}",llm,FormatNotification,ErrorHandler,weather_data,analysis,Analyze this weather data and provide practical advice: {weather_data}. Include temperature, conditions, and helpful recommendations for clothing or activities.
WeatherBot,FormatNotification,,Format the final notification,echo,End,,analysis,notification,
WeatherBot,End,,Weather notification complete,echo,,,notification,final_message,Weather update sent successfully!
WeatherBot,ErrorHandler,,Handle errors gracefully,echo,End,,error,error_message,Unable to get weather data. Please try again later.

📊 Data Processing

🟡 Intermediate

Daily Report Generator

Automated data collection and report generation from multiple sources

Use Case: Generate daily business reports by collecting data from CSV files and creating formatted summaries

reportingautomationdata-aggregation

Required Agents:

csv_readerllmfile_writer

Example Output:

✅ Daily report saved to reports/daily_report.md with executive summary

Configuration Notes:

Ensure CSV files exist in data/ directory
Customize report template in LLM prompts
Add email integration for automatic distribution

View CSV Content

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
DailyReport,LoadSalesData,,"{'format': 'records'}",csv_reader,LoadMetrics,ErrorHandler,collection,sales_data,data/daily_sales.csv
DailyReport,LoadMetrics,,"{'format': 'records'}",csv_reader,AnalyzeData,ErrorHandler,collection,metrics_data,data/metrics.csv
DailyReport,AnalyzeData,,"{'temperature': 0.3, 'model': 'gpt-4'}",llm,GenerateReport,ErrorHandler,sales_data|metrics_data,analysis,Create a comprehensive daily business report from this data: Sales: {sales_data} Metrics: {metrics_data}. Include key insights, trends, and actionable recommendations.
DailyReport,GenerateReport,,"{'mode': 'write'}",file_writer,FormatSummary,ErrorHandler,analysis,report_result,reports/daily_report.md
DailyReport,FormatSummary,,Create executive summary,llm,SaveSummary,ErrorHandler,analysis,summary,Create a 3-bullet executive summary of this report: {analysis}
DailyReport,SaveSummary,,"{'mode': 'write'}",file_writer,End,ErrorHandler,summary,summary_result,reports/executive_summary.txt
DailyReport,End,,Report generation complete,echo,,,summary_result,final_message,Daily report generated successfully!
DailyReport,ErrorHandler,,Handle processing errors,echo,End,,error,error_message,Report generation failed: {error}

🧠 AI/LLM

🟡 Intermediate

Customer Feedback Analyzer

Sentiment analysis and categorization of customer feedback

Use Case: Analyze customer feedback for sentiment, extract key themes, and categorize issues

sentiment-analysiscustomer-servicenlp

Required Agents:

csv_readerllmcsv_writer

Example Output:

📊 Analyzed 150 feedback entries: 68% positive, main themes: shipping delays, product quality

Configuration Notes:

Ensure feedback CSV has 'feedback' and 'customer_id' columns
Adjust sentiment scale in prompts as needed
Add integration with CRM systems for follow-up actions

View CSV Content

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
FeedbackAnalyzer,LoadFeedback,,"{'format': 'records'}",csv_reader,AnalyzeSentiment,ErrorHandler,collection,feedback_data,data/customer_feedback.csv
FeedbackAnalyzer,AnalyzeSentiment,,"{'temperature': 0.2, 'model': 'gpt-4'}",llm,CategorizeIssues,ErrorHandler,feedback_data,sentiment_analysis,Analyze the sentiment of this customer feedback and rate each on a scale of 1-5 (5=very positive, 1=very negative). Return structured data: {feedback_data}
FeedbackAnalyzer,CategorizeIssues,,"{'temperature': 0.3}",llm,ExtractThemes,ErrorHandler,feedback_data|sentiment_analysis,categories,Categorize these customer issues into main themes (e.g., Product Quality, Customer Service, Shipping, etc.): {feedback_data}. Sentiment context: {sentiment_analysis}
FeedbackAnalyzer,ExtractThemes,,Extract key themes and insights,llm,CompileResults,ErrorHandler,feedback_data|sentiment_analysis|categories,themes,Extract the top 3 themes and actionable insights from this feedback analysis: Feedback: {feedback_data}, Sentiment: {sentiment_analysis}, Categories: {categories}
FeedbackAnalyzer,CompileResults,,Compile final analysis,llm,SaveResults,ErrorHandler,sentiment_analysis|categories|themes,final_analysis,Create a comprehensive customer feedback report with: 1) Sentiment summary 2) Issue categories 3) Key themes 4) Recommended actions. Data: Sentiment: {sentiment_analysis}, Categories: {categories}, Themes: {themes}
FeedbackAnalyzer,SaveResults,,"{'format': 'records', 'mode': 'write'}",csv_writer,End,ErrorHandler,final_analysis,save_result,analysis/feedback_analysis.csv
FeedbackAnalyzer,End,,Analysis complete,echo,,,save_result,final_message,Customer feedback analysis saved successfully!
FeedbackAnalyzer,ErrorHandler,,Handle analysis errors,echo,End,,error,error_message,Feedback analysis failed: {error}

👁️ Monitoring

🔴 Advanced

Social Media Monitor

Monitor and analyze social media mentions with alert system

Use Case: Monitor social media mentions, analyze sentiment, and trigger alerts for negative feedback

social-mediamonitoringalertssentiment

Required Agents:

json_readerllmbranchingecho

Example Output:

📱 Monitored 25 mentions: 3 high-priority alerts, 1 urgent response needed

Configuration Notes:

Configure branching logic: negative sentiment (score < 4) AND high influence (> 7) triggers alerts
Integrate with social media APIs for real-time data
Set up webhook notifications for urgent alerts

View CSV Content

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
SocialMonitor,LoadMentions,,"{'format': 'list'}",json_reader,AnalyzeMentions,ErrorHandler,collection,mentions_data,data/social_mentions.json
SocialMonitor,AnalyzeMentions,,"{'temperature': 0.2, 'model': 'gpt-4'}",llm,CheckSentiment,ErrorHandler,mentions_data,analysis,Analyze these social media mentions for sentiment, urgency, and influence level. Return JSON with sentiment_score (1-10), urgency (low/medium/high), and influence_score (1-10): {mentions_data}
SocialMonitor,CheckSentiment,,Check if immediate action needed,branching,TriggerAlert,GenerateReport,analysis,routing_decision,
SocialMonitor,TriggerAlert,,Send immediate alert for negative mentions,echo,GenerateReport,,analysis,alert_sent,🚨 URGENT: Negative social mention detected requiring immediate attention!
SocialMonitor,GenerateReport,,"{'temperature': 0.3}",llm,SaveReport,ErrorHandler,mentions_data|analysis,report,Generate a social media monitoring report including: 1) Mention summary 2) Sentiment trends 3) High-influence accounts 4) Recommended responses. Data: {mentions_data}. Analysis: {analysis}
SocialMonitor,SaveReport,,"{'mode': 'write'}",file_writer,End,ErrorHandler,report,save_result,reports/social_media_report.md
SocialMonitor,End,,Monitoring cycle complete,echo,,,save_result,final_message,Social media monitoring complete. Report saved.
SocialMonitor,ErrorHandler,,Handle monitoring errors,echo,End,,error,error_message,Social media monitoring failed: {error}

📊 Data Processing

🔴 Advanced

Data ETL Pipeline

Extract, transform, and load data between different formats and systems

Use Case: Transform data from CSV to JSON format with validation and enrichment

etldata-transformationcsvjson

Required Agents:

csv_readerllmjson_writercsv_writer

Example Output:

🔄 ETL Complete: 1,250 records transformed, 98.5% data quality score

Configuration Notes:

Customize field mappings in transform prompts
Add data validation rules specific to your use case
Configure output directory structure

View CSV Content

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
DataETL,ExtractData,,"{'format': 'records'}",csv_reader,ValidateData,ErrorHandler,collection,raw_data,data/source_data.csv
DataETL,ValidateData,,"{'temperature': 0.1, 'model': 'gpt-3.5-turbo'}",llm,TransformData,ErrorHandler,raw_data,validation_result,Validate this data for completeness and format. Flag any missing required fields or invalid formats: {raw_data}. Return validation status and list of issues.
DataETL,TransformData,,"{'temperature': 0.2}",llm,EnrichData,ErrorHandler,raw_data|validation_result,transformed_data,Transform this CSV data to JSON format with standardized field names. Apply data cleaning and normalization: {raw_data}. Validation context: {validation_result}
DataETL,EnrichData,,Add calculated fields and metadata,llm,SaveJSON,ErrorHandler,transformed_data,enriched_data,Enrich this data by adding calculated fields, categories, and metadata: {transformed_data}. Add processing timestamp and data quality score.
DataETL,SaveJSON,,"{'format': 'dict', 'indent': 2}",json_writer,CreateSummary,ErrorHandler,enriched_data,json_result,data/output/transformed_data.json
DataETL,CreateSummary,,Generate processing summary,llm,SaveSummary,ErrorHandler,validation_result|json_result,summary,Create an ETL processing summary including: records processed, validation results, transformations applied, and data quality metrics. Validation: {validation_result}, Result: {json_result}
DataETL,SaveSummary,,"{'format': 'records', 'mode': 'write'}",csv_writer,End,ErrorHandler,summary,summary_result,data/output/etl_summary.csv
DataETL,End,,ETL pipeline complete,echo,,,summary_result,final_message,Data ETL pipeline completed successfully!
DataETL,ErrorHandler,,Handle ETL errors,echo,End,,error,error_message,ETL pipeline failed: {error}

🧠 AI/LLM

🟡 Intermediate

Document Summarizer

Intelligent document processing and multi-level summarization

Use Case: Process documents to create executive summaries, key points, and action items

document-processingsummarizationai

Required Agents:

file_readerllmfile_writer

Example Output:

📄 Document processed: 15-page report summarized into 3 key deliverables

Configuration Notes:

Supports PDF, TXT, MD file formats
Adjust chunk_size based on document length
Customize summary style in prompts

View CSV Content

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
DocSummarizer,LoadDocument,,"{'chunk_size': 2000, 'should_split': true}",file_reader,CreateSummary,ErrorHandler,collection,document_content,
DocSummarizer,CreateSummary,,"{'temperature': 0.3, 'model': 'gpt-4'}",llm,ExtractKeyPoints,ErrorHandler,document_content,summary,Create a comprehensive summary of this document. Focus on main themes, key findings, and important conclusions: {document_content}
DocSummarizer,ExtractKeyPoints,,Extract actionable insights,llm,CreateExecutiveSummary,ErrorHandler,document_content|summary,key_points,Extract the 5 most important key points and any action items from this document: {document_content}. Context: {summary}
DocSummarizer,CreateExecutiveSummary,,Create executive-level summary,llm,SaveSummary,ErrorHandler,summary|key_points,executive_summary,Create a concise executive summary (2-3 paragraphs) suitable for leadership review: Summary: {summary}, Key Points: {key_points}
DocSummarizer,SaveSummary,,"{'mode': 'write'}",file_writer,SaveKeyPoints,ErrorHandler,executive_summary,summary_result,output/executive_summary.md
DocSummarizer,SaveKeyPoints,,"{'mode': 'write'}",file_writer,SaveFullAnalysis,ErrorHandler,key_points,keypoints_result,output/key_points.md
DocSummarizer,SaveFullAnalysis,,"{'mode': 'write'}",file_writer,End,ErrorHandler,summary,analysis_result,output/full_analysis.md
DocSummarizer,End,,Document processing complete,echo,,,analysis_result,final_message,Document summarization completed successfully!
DocSummarizer,ErrorHandler,,Handle processing errors,echo,End,,error,error_message,Document processing failed: {error}

👁️ Monitoring

🟡 Intermediate

API Health Checker

Monitor API endpoints and generate health reports with alerting

Use Case: Check API endpoint health, analyze response times, and alert on issues

api-monitoringhealth-checkalerts

Required Agents:

json_readerllmbranchingfile_writer

Example Output:

🔍 Health Check: 12/15 APIs healthy, 2 warnings, 1 critical alert triggered

Configuration Notes:

Configure endpoint list in config/api_endpoints.json
Set alert thresholds in branching logic
Integrate with monitoring tools like Grafana

View CSV Content

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
APIHealthCheck,LoadEndpoints,,"{'format': 'list'}",json_reader,AnalyzeHealth,ErrorHandler,collection,endpoints_data,config/api_endpoints.json
APIHealthCheck,AnalyzeHealth,,"{'temperature': 0.1, 'model': 'gpt-3.5-turbo'}",llm,CheckStatus,ErrorHandler,endpoints_data,health_analysis,Analyze this API health data and determine status for each endpoint. Look for response times >500ms, error rates >1%, and downtime. Return structured status: {endpoints_data}
APIHealthCheck,CheckStatus,,Determine if alerts needed,branching,TriggerAlert,GenerateReport,health_analysis,alert_decision,
APIHealthCheck,TriggerAlert,,Send critical alerts,echo,GenerateReport,,health_analysis,alert_sent,🚨 API ALERT: Critical endpoints detected requiring immediate attention!
APIHealthCheck,GenerateReport,,"{'temperature': 0.2}",llm,SaveReport,ErrorHandler,endpoints_data|health_analysis,health_report,Generate a comprehensive API health report including: 1) Endpoint status summary 2) Performance metrics 3) Error analysis 4) Recommendations. Data: {endpoints_data}. Analysis: {health_analysis}
APIHealthCheck,SaveReport,,"{'mode': 'write'}",file_writer,CreateDashboard,ErrorHandler,health_report,report_result,reports/api_health_report.md
APIHealthCheck,CreateDashboard,,Create monitoring dashboard data,llm,SaveDashboard,ErrorHandler,health_analysis,dashboard_data,Create dashboard-ready JSON data from this health analysis for visualization: {health_analysis}
APIHealthCheck,SaveDashboard,,"{'mode': 'write'}",file_writer,End,ErrorHandler,dashboard_data,dashboard_result,monitoring/dashboard_data.json
APIHealthCheck,End,,Health check complete,echo,,,dashboard_result,final_message,API health check completed. Reports generated.
APIHealthCheck,ErrorHandler,,Handle monitoring errors,echo,End,,error,error_message,API health check failed: {error}

🧠 AI/LLM

🟢 Beginner

Email Classifier

Intelligent email categorization and priority routing system

Use Case: Automatically categorize and prioritize incoming emails based on content and sender

email-processingclassificationautomation

Required Agents:

csv_readerllmbranchingcsv_writer

Example Output:

📧 Processed 45 emails: 3 urgent, 15 support tickets, 12 sales inquiries

Configuration Notes:

Ensure email CSV has 'subject', 'sender', 'content' columns
Configure urgency keywords for branching logic
Add integration with email systems for automation

View CSV Content

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
EmailClassifier,LoadEmails,,"{'format': 'records'}",csv_reader,ClassifyEmails,ErrorHandler,collection,email_data,data/incoming_emails.csv
EmailClassifier,ClassifyEmails,,"{'temperature': 0.2, 'model': 'gpt-3.5-turbo'}",llm,DeterminePriority,ErrorHandler,email_data,classification,Classify these emails into categories: Support, Sales, Marketing, Technical, Urgent. Also determine sentiment (positive/neutral/negative): {email_data}
EmailClassifier,DeterminePriority,,Assess priority level,llm,RouteEmails,ErrorHandler,email_data|classification,priority_assessment,Determine priority level (High/Medium/Low) for each email based on content urgency, sender importance, and keywords: {email_data}. Classification context: {classification}
EmailClassifier,RouteEmails,,Route based on classification,branching,ProcessUrgent,ProcessNormal,priority_assessment,routing_decision,
EmailClassifier,ProcessUrgent,,Handle urgent emails,echo,SaveResults,,priority_assessment,urgent_processed,⚡ Urgent emails flagged for immediate attention
EmailClassifier,ProcessNormal,,Handle normal priority emails,echo,SaveResults,,priority_assessment,normal_processed,📧 Standard emails categorized and queued
EmailClassifier,SaveResults,,"{'format': 'records', 'mode': 'write'}",csv_writer,GenerateSummary,ErrorHandler,classification|priority_assessment,save_result,data/classified_emails.csv
EmailClassifier,GenerateSummary,,Create processing summary,llm,End,ErrorHandler,classification|priority_assessment,summary,Create an email processing summary with category counts and priority distribution: Classifications: {classification}, Priorities: {priority_assessment}
EmailClassifier,End,,Email classification complete,echo,,,summary,final_message,Email classification completed successfully!
EmailClassifier,ErrorHandler,,Handle classification errors,echo,End,,error,error_message,Email classification failed: {error}

🧠 AI/LLM

🟡 Intermediate

Translation Workflow

Multi-language document translation with quality assurance

Use Case: Translate documents between languages with quality checks and formatting preservation

translationmultilingualqa

Required Agents:

file_readerllmfile_writer

Example Output:

🌐 Translation complete: EN→ES, Quality score: 9.2/10, 3 pages processed

Configuration Notes:

Specify target language in initial input
Adjust chunk_size for optimal translation context
Add glossary terms for domain-specific translation

View CSV Content

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
TranslationWorkflow,LoadDocument,,"{'chunk_size': 1500, 'should_split': true}",file_reader,DetectLanguage,ErrorHandler,collection,source_content,
TranslationWorkflow,DetectLanguage,,"{'temperature': 0.1}",llm,TranslateContent,ErrorHandler,source_content,language_info,Detect the source language of this text and confirm the target language for translation: {source_content}. Respond with source_language and confidence_level.
TranslationWorkflow,TranslateContent,,"{'temperature': 0.3, 'model': 'gpt-4'}",llm,QualityCheck,ErrorHandler,source_content|language_info,translation,Translate this text from the detected source language to the target language. Preserve formatting, maintain professional tone, and ensure cultural appropriateness: {source_content}. Language context: {language_info}
TranslationWorkflow,QualityCheck,,Review translation quality,llm,FormatOutput,ErrorHandler,source_content|translation,quality_review,Review this translation for accuracy, fluency, and completeness. Rate quality (1-10) and note any issues: Original: {source_content}. Translation: {translation}
TranslationWorkflow,FormatOutput,,Format final translation,llm,SaveTranslation,ErrorHandler,translation|quality_review,formatted_translation,Format the final translation with proper structure and any necessary corrections: {translation}. Quality notes: {quality_review}
TranslationWorkflow,SaveTranslation,,"{'mode': 'write'}",file_writer,SaveQualityReport,ErrorHandler,formatted_translation,translation_result,output/translated_document.txt
TranslationWorkflow,SaveQualityReport,,"{'mode': 'write'}",file_writer,End,ErrorHandler,quality_review,quality_result,output/translation_quality_report.txt
TranslationWorkflow,End,,Translation workflow complete,echo,,,quality_result,final_message,Document translation completed with quality report!
TranslationWorkflow,ErrorHandler,,Handle translation errors,echo,End,,error,error_message,Translation workflow failed: {error}

🧠 AI/LLM

🔴 Advanced

Content Moderator

AI-powered content moderation with policy compliance checking

Use Case: Moderate user-generated content for safety, policy compliance, and quality standards

content-moderationsafetycompliance

Required Agents:

csv_readerllmbranchingcsv_writer

Example Output:

🛡️ Moderated 200 posts: 185 approved, 12 flagged for review, 3 removed

Configuration Notes:

Configure policy rules in branching conditions
Adjust safety thresholds based on platform needs
Add human review queue for borderline cases

View CSV Content

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
ContentModerator,LoadContent,,"{'format': 'records'}",csv_reader,InitialScreen,ErrorHandler,collection,content_data,data/user_content.csv
ContentModerator,InitialScreen,,"{'temperature': 0.1, 'model': 'gpt-4'}",llm,DeepAnalysis,ErrorHandler,content_data,initial_screening,Screen this content for obvious policy violations, inappropriate language, and spam. Rate safety level (1-10, 10=completely safe): {content_data}
ContentModerator,DeepAnalysis,,Detailed content analysis,llm,PolicyCheck,ErrorHandler,content_data|initial_screening,detailed_analysis,Perform detailed analysis of this content for: 1) Hate speech 2) Violence 3) Sexual content 4) Harassment 5) Misinformation. Content: {content_data}. Initial screening: {initial_screening}
ContentModerator,PolicyCheck,,Check against content policies,llm,DetermineAction,ErrorHandler,content_data|detailed_analysis,policy_compliance,Check this content against platform policies and community guidelines. Determine if content should be: approved, flagged for review, or removed. Analysis: {detailed_analysis}
ContentModerator,DetermineAction,,Decide on moderation action,branching,FlagContent,ApproveContent,policy_compliance,moderation_decision,
ContentModerator,FlagContent,,Flag problematic content,echo,SaveResults,,policy_compliance,flagged_result,🚩 Content flagged for manual review or removal
ContentModerator,ApproveContent,,Approve safe content,echo,SaveResults,,policy_compliance,approved_result,✅ Content approved for publication
ContentModerator,SaveResults,,"{'format': 'records', 'mode': 'write'}",csv_writer,GenerateReport,ErrorHandler,initial_screening|detailed_analysis|policy_compliance,save_result,data/moderation_results.csv
ContentModerator,GenerateReport,,Create moderation summary,llm,End,ErrorHandler,initial_screening|detailed_analysis|policy_compliance,moderation_report,Generate a content moderation report with statistics, flagged items, and trend analysis: Screening: {initial_screening}, Analysis: {detailed_analysis}, Compliance: {policy_compliance}
ContentModerator,End,,Content moderation complete,echo,,,moderation_report,final_message,Content moderation completed successfully!
ContentModerator,ErrorHandler,,Handle moderation errors,echo,End,,error,error_message,Content moderation failed: {error}

Found 10 of 10 templates. Need help customizing a template? Check out the template customization guide or quick start tutorial.

How to Use Templates

1. Choose Your Template

Browse the template library above to find a workflow that matches your needs. Use the category and difficulty filters to narrow down your options.

2. Copy the CSV Content

Click the "📋 Copy CSV" button on any template to copy the workflow definition to your clipboard.

3. Load into AgentMap

Paste the CSV content into AgentMap using one of these methods:

Command Line: Save as a .csv file and run:
```
agentmap execute your_workflow.csv
```

Python API: Load directly in your code:

from agentmap import AgentMap

# Paste CSV content here
csv_content = """GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
..."""

agent_map = AgentMap()
result = agent_map.execute_from_csv_string(csv_content)

Playground: Use the "🚀 Open in Playground" button to launch the template directly in the AgentMap web interface.

4. Customize Configuration

Most templates include configuration notes with specific customization instructions. Common customizations include:

File Paths: Update input/output paths for your directory structure
API Keys: Configure LLM providers and external services
Prompts: Modify prompts to match your specific domain or tone
Agent Parameters: Adjust temperature, model selection, and other agent settings

Template Categories

🤖 Automation

Workflows that automate repetitive tasks and processes:

Weather Notification Bot: Daily weather alerts with intelligent notifications
Email Classifier: Automatic email categorization and priority routing

📊 Data Processing

Templates for data transformation, analysis, and reporting:

Daily Report Generator: Automated data collection and report generation
Data ETL Pipeline: Extract, transform, and load data between systems

🧠 AI/LLM

AI-powered workflows leveraging language models:

Customer Feedback Analyzer: Sentiment analysis and issue categorization
Document Summarizer: Multi-level document processing and summarization
Translation Workflow: Multi-language translation with quality assurance
Content Moderator: AI-powered content moderation and compliance

👁️ Monitoring

Real-time monitoring and alerting systems:

Social Media Monitor: Track mentions with sentiment analysis and alerts
API Health Checker: Monitor endpoint health with automated reporting

🔗 Integration

Templates for connecting different systems and services:

Data ETL Pipeline: Seamless data movement between formats

🛠️ Utility

General-purpose workflows for common tasks:

Various utility templates for file processing, data validation, and more

Difficulty Levels

🟢 Beginner

Perfect for new users learning AgentMap:

Simple, linear workflows
Basic agent types (echo, input, llm)
Minimal configuration required
Clear documentation and examples

🟡 Intermediate

For users comfortable with AgentMap basics:

Multi-step workflows with branching
Multiple agent types and data formats
Some external integrations
Customizable parameters

🔴 Advanced

Complex workflows for experienced users:

Sophisticated routing and orchestration
Multiple data sources and outputs
External API integrations
Advanced error handling

Customization Guide

Modifying Prompts

LLM agent prompts can be customized to match your specific needs:

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
MyWorkflow,Analyzer,,"{'temperature': 0.3}",llm,Next,,input,analysis,"Analyze this data for trends and insights: {input}. Focus on actionable recommendations."

Tips for prompt customization:

Be specific about the desired output format
Include examples of good responses
Set the appropriate tone (formal, casual, technical)
Use field placeholders like {input} for dynamic content

Configuring Agent Context

Many agents accept context parameters for fine-tuning:

Context
"{'temperature': 0.7, 'model': 'gpt-4', 'max_tokens': 500}"
"{'format': 'records', 'encoding': 'utf-8'}"
"{'chunk_size': 1000, 'should_split': true}"

Common context options:

LLM agents: temperature, model, max_tokens
File agents: encoding, mode, chunk_size
CSV agents: format, delimiter, id_field

Adding Error Handling

Robust workflows include proper error handling:

GraphName,Node,Edge,Context,AgentType,Success_Next,Failure_Next,Input_Fields,Output_Field,Prompt
MyWorkflow,ProcessData,,Process the data,llm,SaveResult,ErrorHandler,input,result,"Process: {input}"
MyWorkflow,ErrorHandler,,Handle errors gracefully,echo,End,,error,error_msg,"Error occurred: {error}"

File Path Configuration

Update file paths to match your directory structure:

# Input files
"data/input.csv"
"config/settings.json"

# Output files  
"reports/daily_summary.md"
"output/processed_data.csv"

Best Practices

1. Start Simple

Begin with beginner templates and gradually work up to more complex workflows as you become comfortable with AgentMap concepts.

2. Test Incrementally

When customizing templates:

Make small changes at a time
Test each modification before adding more
Use the error messages to guide troubleshooting

3. Organize Your Files

Create a clear directory structure for your workflows:

my_agentmap_project/
├── workflows/
│   ├── daily_reports.csv
│   ├── content_moderation.csv
│   └── data_processing.csv
├── data/
│   ├── input/
│   └── output/
├── config/
│   └── settings.json
└── reports/

4. Version Control

Keep your customized workflows in version control to track changes and collaborate with team members.

5. Document Customizations

When modifying templates, document your changes:

What was changed and why
Any new dependencies or requirements
Expected input/output formats

Troubleshooting

Common Issues

CSV Format Errors

Ensure all rows have the same number of columns
Check for unescaped commas in text fields
Verify column headers match expected format

Agent Configuration

Validate JSON syntax in Context fields
Check that required agent types are available
Ensure file paths exist and are accessible

Missing Dependencies

Install required Python packages
Configure API keys for external services
Verify file permissions for input/output directories

Getting Help

Need assistance with templates?

Documentation: Check the Agent Types Reference for detailed agent documentation
Quick Start: Review the Quick Start Guide for AgentMap basics
Examples: Explore additional examples in the Examples Directory
Community: Join discussions in our community forums

Contributing Templates

Have a useful workflow template to share? We welcome contributions!

Template Requirements

Well-documented use case and configuration
Tested and working example
Clear setup instructions
Appropriate difficulty classification

Submission Process

Create your template following our format
Test thoroughly with sample data
Document configuration requirements
Submit via pull request with description

See our Contributing Guide for detailed submission instructions.

Ready to get started? Choose a template above that matches your use case and start building your first AgentMap workflow!

AgentMap Template Library

Weather Notification Bot

Daily Report Generator

Customer Feedback Analyzer

Social Media Monitor

Data ETL Pipeline

Document Summarizer

API Health Checker

Email Classifier

Translation Workflow

Content Moderator

How to Use Templates​

1. Choose Your Template​

2. Copy the CSV Content​

3. Load into AgentMap​

4. Customize Configuration​

Template Categories​

🤖 Automation​

📊 Data Processing​

🧠 AI/LLM​

👁️ Monitoring​

🔗 Integration​

🛠️ Utility​

Difficulty Levels​

🟢 Beginner​

🟡 Intermediate​

🔴 Advanced​

Customization Guide​

Modifying Prompts​

Configuring Agent Context​

Adding Error Handling​

File Path Configuration​

Best Practices​

1. Start Simple​

2. Test Incrementally​

3. Organize Your Files​

4. Version Control​

5. Document Customizations​

Troubleshooting​

Common Issues​

Getting Help​

Contributing Templates​

Template Requirements​

Submission Process​

How to Use Templates

1. Choose Your Template

2. Copy the CSV Content

3. Load into AgentMap

4. Customize Configuration

Template Categories

🤖 Automation

📊 Data Processing

🧠 AI/LLM

👁️ Monitoring

🔗 Integration

🛠️ Utility

Difficulty Levels

🟢 Beginner

🟡 Intermediate

🔴 Advanced

Customization Guide

Modifying Prompts

Configuring Agent Context

Adding Error Handling

File Path Configuration

Best Practices

1. Start Simple

2. Test Incrementally

3. Organize Your Files

4. Version Control

5. Document Customizations

Troubleshooting

Common Issues

Getting Help

Contributing Templates

Template Requirements

Submission Process