Skip to content

EXAMPLES.md

Latest commit

 

History

History
522 lines (358 loc) · 9.02 KB

EXAMPLES.md

File metadata and controls

522 lines (358 loc) · 9.02 KB
Raw

Examples

This page provides an index of all example applications included with Minion.

Quick Start

# Clone the repository
git clone https://github.com/Ranganaths/minion.git
cd minion

# Set API key
export OPENAI_API_KEY="sk-..."

# Run an example
go run examples/basic/main.go

Basic Examples

Basic Agent

Create and execute a simple agent with LLM.

go run examples/basic/main.go

Demonstrates:

  • Framework initialization
  • Agent creation and activation
  • Basic execution
  • Metrics retrieval

Location: examples/basic/

Agent with Tools

Agent that can use built-in tools.

go run examples/with_tools/main.go

Demonstrates:

  • Tool registration
  • Capability-based tool filtering
  • Tool execution
  • Tool output handling

Location: examples/with_tools/

Custom Behavior

Create a custom behavior for specialized agent processing.

go run examples/custom_behavior/main.go

Demonstrates:

  • Behavior interface implementation
  • Custom system prompts
  • Input/output processing
  • Behavior registration

Location: examples/custom_behavior/

Domain Tools

Use domain-specific tools for various tasks.

go run examples/domain_tools/main.go

Demonstrates:

  • Data tools (CSV, JSON, XML)
  • HTTP tools
  • File operations
  • Security tools

Location: examples/domain_tools/

Multi-Agent Examples

Multi-Agent Basic

Simple multi-agent system with coordinator and workers.

go run examples/multiagent-basic/main.go

Demonstrates:

  • Orchestrator creation
  • Worker specialization
  • Task delegation
  • Result aggregation

Location: examples/multiagent-basic/

Multi-Agent Custom

Advanced multi-agent patterns with custom workflows.

go run examples/multiagent-custom/main.go

Demonstrates:

  • Custom worker implementations
  • Complex task decomposition
  • Inter-agent communication
  • Error handling

Location: examples/multiagent-custom/

Multi-Agent Tracing

Multi-agent system with distributed tracing.

go run examples/multiagent-tracing/main.go

Demonstrates:

  • Distributed traces across agents
  • Span propagation
  • Performance monitoring
  • Trace visualization

Location: examples/multiagent-tracing/

Protocol Examples

A2A Server

Expose agent via A2A protocol for inter-agent communication.

go run examples/a2a-server/main.go

Demonstrates:

  • A2A server setup
  • Agent Card generation
  • JSON-RPC endpoints
  • SSE streaming

Endpoints:

  • GET /.well-known/agent.json - Agent Card
  • POST / - JSON-RPC endpoint

Location: examples/a2a-server/

AG-UI Server

Agent with real-time streaming for frontend applications.

go run examples/agui-server/main.go

Demonstrates:

  • AG-UI server setup
  • SSE event streaming
  • Token-by-token output
  • State synchronization

Endpoints:

  • POST / - Run request with SSE

Location: examples/agui-server/

Industry Examples

Sales Agent

AI sales assistant with CRM integration.

go run examples/sales_agent/main.go

Demonstrates:

  • Lead qualification
  • CRM tool integration
  • Email composition
  • Sales pipeline management

Location: examples/sales_agent/

Sales Automation

Full sales workflow automation.

go run examples/sales-automation/main.go

Demonstrates:

  • Lead scoring
  • Automated outreach
  • Follow-up scheduling
  • Analytics

Location: examples/sales-automation/

Customer Support

AI-powered customer support agent.

go run examples/customer-support/main.go

Demonstrates:

  • Ticket classification
  • Knowledge base search
  • Response generation
  • Escalation handling

Location: examples/customer-support/

Business Automation

General business process automation.

go run examples/business_automation/main.go

Demonstrates:

  • Document processing
  • Approval workflows
  • Report generation
  • Integration patterns

Location: examples/business_automation/

DevOps Automation

Infrastructure and deployment automation.

go run examples/devops-automation/main.go

Demonstrates:

  • CI/CD integration
  • Infrastructure management
  • Monitoring automation
  • Incident response

Location: examples/devops-automation/

Advanced Examples

Chain Features

LLM chain patterns including RAG, sequential, and router chains.

go run examples/chain-features/main.go

Demonstrates:

  • LLM chains
  • RAG chains
  • Sequential chains
  • Router chains
  • Transform chains

Location: examples/chain-features/

LLM Worker

Dedicated LLM worker pattern.

go run examples/llm_worker/main.go

Demonstrates:

  • Worker pool pattern
  • Request queuing
  • Batch processing
  • Rate limiting

Location: examples/llm_worker/

TupleLeap Integration

Using TupleLeap as LLM provider.

go run examples/tupleleap_example/main.go

Demonstrates:

  • TupleLeap provider setup
  • Custom model configuration
  • API integration

Location: examples/tupleleap_example/

Observability Examples

Tracing

OpenTelemetry distributed tracing.

go run examples/tracing/main.go

Demonstrates:

  • Tracer initialization
  • Span creation
  • Attribute recording
  • Error tracking

Location: examples/tracing/

Full Observability

Complete observability setup with tracing and metrics.

go run examples/observability/main.go

Demonstrates:

  • OpenTelemetry tracing
  • Prometheus metrics
  • Custom instrumentation
  • Dashboard integration

Location: examples/observability/

Debug Time-Travel

Time-travel debugging for agent execution.

go run examples/debug-timetravel/main.go

Demonstrates:

  • Execution recording
  • State snapshots
  • Timeline navigation
  • Replay debugging

Location: examples/debug-timetravel/

Evaluation Examples

Agent Evaluation

Benchmark and evaluate agent performance.

go run examples/evaluation/main.go

Demonstrates:

  • Test case creation
  • Evaluation metrics
  • Performance benchmarks
  • Quality assessment

Location: examples/evaluation/

Self-Improving Agent

Agent that improves based on feedback.

go run examples/self-improving/main.go

Demonstrates:

  • Feedback collection
  • Performance analysis
  • Prompt optimization
  • Iterative improvement

Location: examples/self-improving/

Running Examples

Prerequisites

  1. Go 1.24+ installed
  2. API key set as environment variable
# OpenAI
export OPENAI_API_KEY="sk-..."

# Or Anthropic
export ANTHROPIC_API_KEY="sk-ant-..."

# Or TupleLeap
export TUPLELEAP_API_KEY="..."

Running

# From project root
go run examples/<example-name>/main.go

# Or build first
go build -o example examples/<example-name>/main.go
./example

Common Issues

"API key not set"

export OPENAI_API_KEY="your-key-here"

"Module not found"

go mod tidy

"Port already in use" (for server examples)

# Use different port
PORT=9090 go run examples/a2a-server/main.go

Creating Your Own Examples

Use the basic example as a template:

package main

import (
    "context"
    "fmt"
    "os"

    "github.com/Ranganaths/minion/core"
    "github.com/Ranganaths/minion/llm"
    "github.com/Ranganaths/minion/models"
    "github.com/Ranganaths/minion/storage"
)

func main() {
    // 1. Create framework
    framework := core.NewFramework(
        core.WithStorage(storage.NewInMemory()),
        core.WithLLMProvider(llm.NewOpenAI(os.Getenv("OPENAI_API_KEY"))),
    )
    defer framework.Close()

    // 2. Create agent
    ctx := context.Background()
    agent, _ := framework.CreateAgent(ctx, &models.CreateAgentRequest{
        Name:        "My Example Agent",
        Description: "Description of what this agent does",
    })

    // 3. Activate
    activeStatus := models.StatusActive
    framework.UpdateAgent(ctx, agent.ID, &models.UpdateAgentRequest{
        Status: &activeStatus,
    })

    // 4. Execute
    output, _ := framework.Execute(ctx, agent.ID, &models.Input{
        Raw: "Your prompt here",
    })

    fmt.Println(output.Result)
}

Next Steps