SwarmGo

https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo
Admin
SwarmGo is a Go package for creating AI agents capable of interacting, coordinating, and executing...
Tokens:24,644
Snippets:315
Trust Score:9.6
Update:5 months ago
Show doc for...
Tokens:
Raw
### Quick Start: Run a Basic Agent

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

This example demonstrates how to initialize a Swarm client, create a simple agent, and run it with an initial message. Ensure you replace 'YOUR_OPENAI_API_KEY' with your actual OpenAI API key.

```go
package main

import (
	"context"
	"fmt"
	"log"

	swarmgo "github.com/prathyushnallamothu/swarmgo"
	openai "github.com/sashabaranov/go-openai"
	llm "github.com/prathyushnallamothu/swarmgo/llm"
)

func main() {
	client := swarmgo.NewSwarm("YOUR_OPENAI_API_KEY", llm.OpenAI)

	agent := &swarmgo.Agent{
		Name:         "Agent",
		Instructions: "You are a helpful assistant.",
		Model:        "gpt-3.5-turbo",
	}

	messages := []openai.ChatCompletionMessage{
		{Role: "user", Content: "Hello!"},
	}

	ctx := context.Background()
	response, err := client.Run(ctx, agent, messages, nil, "", false, false, 5, true)
	if err != nil {
		log.Fatalf("Error: %v", err)
	}

	fmt.Println(response.Messages[len(response.Messages)-1].Content)
}

```

--------------------------------

### Install SwarmGo Package

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Use 'go get' to install the SwarmGo package. This command fetches and installs the specified package and its dependencies.

```bash
go get github.com/prathyushnallamothu/swarmgo

```

--------------------------------

### Run Demo Loop with Custom Configuration

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Starts an interactive CLI loop with custom configuration options.

```go
func RunDemoLoopWithConfig(client *Swarm, agent *Agent, config *DemoLoopConfig)
{
	// ... implementation details ...
}
```

--------------------------------

### Run Interactive Demo Loop

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Starts an interactive command-line interface (CLI) loop for conversing with an agent.

```go
func RunDemoLoop(client *Swarm, agent *Agent)
{
	// ... implementation details ...
}
```

--------------------------------

### Execute Graph

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Executes the workflow graph starting from its entry point with an initial state.

```go
func (g *Graph) ExecuteGraph(ctx context.Context, initialState GraphState) (GraphState, error)
```

--------------------------------

### Set Graph Entry Point

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Specifies the starting node for the workflow graph execution.

```go
func (g *Graph) SetEntryPoint(nodeID NodeID) error
```

--------------------------------

### Get All Step Results

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Returns all step results from the workflow execution. Added in v1.0.4.

```go
func (wf *Workflow) GetAllStepResults() []StepResult
```

--------------------------------

### Create Hierarchical Workflow in SwarmGo

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Illustrates the setup of a Hierarchical workflow, assigning agents to teams (manager, research, analysis), and establishing the agent connection hierarchy.

```go
workflow := swarmgo.NewWorkflow(apiKey, llm.OpenAI, swarmgo.HierarchicalWorkflow)

// Add agents to teams
workflow.AddAgentToTeam(managerAgent, swarmgo.SupervisorTeam)
workflow.AddAgentToTeam(researchAgent, swarmgo.ResearchTeam)
workflow.AddAgentToTeam(analysisAgent, swarmgo.AnalysisTeam)

// Connect agents in hierarchy
workflow.ConnectAgents(managerAgent.Name, researchAgent.Name)
workflow.ConnectAgents(researchAgent.Name, analysisAgent.Name)


```

--------------------------------

### Implement Custom StreamHandler

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Create a custom StreamHandler to define specific behavior for streaming events. This example tracks total tokens and prints output.

```go
type CustomStreamHandler struct {
    totalTokens int
}

func (h *CustomStreamHandler) OnStart() {
    fmt.Println("Starting stream...")
}

func (h *CustomStreamHandler) OnToken(token string) {
    h.totalTokens++
    fmt.Print(token)
}

func (h *CustomStreamHandler) OnComplete(msg openai.ChatCompletionMessage) {
    fmt.Printf("\nComplete! Total tokens: %d\n", h.totalTokens)
}

func (h *CustomStreamHandler) OnError(err error) {
    fmt.Printf("Error: %v\n", err)
}

func (h *CustomStreamHandler) OnToolCall(tool openai.ToolCall) {
    fmt.Printf("\nUsing tool: %s\n", tool.Function.Name)
}
```

--------------------------------

### Execute Workflow

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Runs the workflow starting from a specified agent with a user request, returning detailed results. Added in v1.0.4.

```go
func (wf *Workflow) Execute(startAgent string, userRequest string) (*WorkflowResult, error)
```

--------------------------------

### Get All Agents

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Retrieves all agents currently in the workflow. Added in v1.0.4.

```go
func (wf *Workflow) GetAgents() map[string]*Agent
```

--------------------------------

### Agent.WithInstructions

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Sets static instructions for the agent, guiding its responses and behavior.

```APIDOC
## Agent.WithInstructions

### Description
Sets the static instructions for the agent.

### Signature
```go
func (a *Agent) WithInstructions(instructions string) *Agent
```
```

--------------------------------

### RunDemoLoop

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Starts an interactive command-line interface (CLI) loop for conversing with a specified agent.

```APIDOC
## RunDemoLoop

### Description
RunDemoLoop starts an interactive CLI loop for conversing with an agent.

### Signature
```go
func RunDemoLoop(client *Swarm, agent *Agent)
```
```

--------------------------------

### Get Workflow Connections

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Returns all connections between agents in the workflow. Added in v1.0.4.

```go
func (wf *Workflow) GetConnections() map[string][]string
```

--------------------------------

### Get Routing Log

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Returns the routing history of the workflow. Added in v1.0.4.

```go
func (wf *Workflow) GetRoutingLog() []string
```

--------------------------------

### Get Teams and Agents

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Returns a map of team types to slices of agents belonging to each team. Added in v1.0.4.

```go
func (wf *Workflow) GetTeams() map[TeamType][]*Agent
```

--------------------------------

### Get Default Demo Loop Configuration

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

DefaultDemoLoopConfig returns a pointer to a DemoLoopConfig struct initialized with default values for the demo loop.

```go
func DefaultDemoLoopConfig() *DemoLoopConfig
```

--------------------------------

### Execute Workflow Graph with GraphRunner

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Executes a registered workflow graph identified by graphID, starting with the provided initialState. Returns the final state and any errors encountered.

```go
func (r *GraphRunner) ExecuteGraph(ctx context.Context, graphID string, initialState GraphState) (GraphState, error)
```

--------------------------------

### Execute

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Executes the workflow starting from a specified agent, processing a user's request. Returns the workflow result, including detailed step outcomes, or an error if execution fails. Added in v1.0.4.

```APIDOC
## Execute

### Description
Execute runs the workflow and returns detailed results including step outcomes.

### Method
func (*Workflow) Execute(startAgent string, userRequest string) (*WorkflowResult, error)
```

--------------------------------

### Initialize SwarmGo with LLM Provider

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Shows how to initialize a SwarmGo client with either OpenAI or Gemini as the language learning model provider.

```go
// Initialize with OpenAI
client := swarmgo.NewSwarm("YOUR_API_KEY", llm.OpenAI)

```

```go
// Initialize with Gemini
client := swarmgo.NewSwarm("YOUR_API_KEY", llm.Gemini)

```

--------------------------------

### Initialize Swarm with Configuration

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Initializes a new Swarm with custom configuration. Requires API key, LLM provider, and a Config pointer.

```go
func NewSwarmWithConfig(apiKey string, provider llm.LLMProvider, config *Config) *Swarm
```

--------------------------------

### Initialize Swarm with Host

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Creates a Swarm instance with a custom host. Requires API key, host string, and an LLM provider.

```go
func NewSwarmWithHost(apiKey, host string, provider llm.LLMProvider) *Swarm
```

--------------------------------

### Get Recent Memories from MemoryStore

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Retrieves the 'n' most recently added memories from the MemoryStore.

```go
func (ms *MemoryStore) GetRecentMemories(n int) []Memory
```

--------------------------------

### Configure Agent with Static Instructions

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Sets the static instructions for an agent.

```go
func (a *Agent) WithInstructions(instructions string) *Agent
```

--------------------------------

### Run

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

The main entry point for agent execution.

```APIDOC
## Run

### Description
Run is the main entry point for agent execution.

### Signature
```go
func (s *Swarm) Run(
	ctx context.Context,
	agent *Agent,
	messages []llm.Message,
	contextVariables map[string]interface{},
	modelOverride string,
	stream bool,
	debug bool,
	maxTurns int,
	executeTools bool,
) (Response, error)
```
```

--------------------------------

### Create an Agent Configuration

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Defines the basic structure of an agent, including its name, instructions (system prompt), and the OpenAI model to be used.

```go
agent := &swarmgo.Agent{
	Name:         "Agent",
	Instructions: "You are a helpful assistant.",
	Model:        "gpt-4o",
}

```

--------------------------------

### Initialize and Configure Collaborative Workflow

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Set up a new collaborative workflow with a specified API key and language model. Agents are then added to a 'DocumentTeam' and connected in a circular pattern for collaborative task processing.

```go
workflow := swarmgo.NewWorkflow(apiKey, llm.OpenAI, swarmgo.CollaborativeWorkflow)

// Add agents to document team
workflow.AddAgentToTeam(editor, swarmgo.DocumentTeam)
workflow.AddAgentToTeam(reviewer, swarmgo.DocumentTeam)
workflow.AddAgentToTeam(writer, swarmgo.DocumentTeam)

// Connect agents in collaborative pattern
workflow.ConnectAgents(editor.Name, reviewer.Name)
workflow.ConnectAgents(reviewer.Name, writer.Name)
workflow.ConnectAgents(writer.Name, editor.Name)
```

--------------------------------

### Get Current Agent

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Returns the name of the currently active agent in the workflow. Added in v1.0.4.

```go
func (wf *Workflow) GetCurrentAgent() string
```

--------------------------------

### NewSwarmWithConfig

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Initializes a new Swarm with custom configuration.

```APIDOC
## NewSwarmWithConfig

### Description
Initializes a new Swarm with custom configuration.

### Signature
```go
func NewSwarmWithConfig(apiKey string, provider llm.LLMProvider, config *Config) *Swarm
```
```

--------------------------------

### Get Specific Step Result

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Returns the result of a specific step in the workflow by its number. Added in v1.0.4.

```go
func (wf *Workflow) GetStepResult(stepNumber int) (*StepResult, error)
```

--------------------------------

### Create New GraphBuilder

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Initializes a new GraphBuilder with a name and description for the graph being built.

```go
func NewGraphBuilder(name, description string) *GraphBuilder
```

--------------------------------

### Get Last Step Result

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Returns the result of the last executed step in the workflow. Added in v1.0.4.

```go
func (wf *Workflow) GetLastStepResult() (*StepResult, error)
```

--------------------------------

### Create New Graph

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Initializes a new workflow graph with a given name and description.

```go
func NewGraph(name string, description string) *Graph
```

--------------------------------

### Configure Agent with ClientConfig

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Sets the client configuration for an agent.

```go
func (a *Agent) WithConfig(config *ClientConfig) *Agent
```

--------------------------------

### Create Supervisor Workflow in SwarmGo

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Demonstrates setting up a Supervisor workflow, adding agents to specific teams (supervisor and worker), designating a team leader, and connecting agents for communication.

```go
workflow := swarmgo.NewWorkflow(apiKey, llm.OpenAI, swarmgo.SupervisorWorkflow)

// Add agents to teams
workflow.AddAgentToTeam(supervisorAgent, swarmgo.SupervisorTeam)
workflow.AddAgentToTeam(workerAgent1, swarmgo.WorkerTeam)
workflow.AddAgentToTeam(workerAgent2, swarmgo.WorkerTeam)

// Set supervisor as team leader
workflow.SetTeamLeader(supervisorAgent.Name, swarmgo.SupervisorTeam)

// Connect agents
workflow.ConnectAgents(supervisorAgent.Name, workerAgent1.Name)
workflow.ConnectAgents(supervisorAgent.Name, workerAgent2.Name)


```

--------------------------------

### Stream Handler OnStart Method

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

OnStart is a method of DefaultStreamHandler called when a stream operation begins. It takes no arguments.

```go
func (h *DefaultStreamHandler) OnStart()
```

--------------------------------

### Get Team Leaders

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Returns a map of team types to their respective leader agent names. Added in v1.0.4.

```go
func (wf *Workflow) GetTeamLeaders() map[TeamType]string
```

--------------------------------

### Get Value from GraphState

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Retrieves a value from the GraphState using a StateKey. Includes type assertion to safely access the value.

```go
func (s GraphState) Get(key StateKey) interface{}
```

--------------------------------

### Define Agent Instructions with Context Variables

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Use context variables to dynamically set agent instructions. Ensure type assertion for context variables.

```go
func instructions(contextVariables map[string]interface{}) string {
	name, ok := contextVariables["name"].(string)
	if !ok {
		name = "User"
	}
	return fmt.Sprintf("You are a helpful assistant. Greet the user by name (%s).", name)
}

agent.InstructionsFunc = instructions
```

--------------------------------

### Get Default Swarm Configuration

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

DefaultConfig returns a pointer to a Config struct initialized with default Swarm configuration values.

```go
func DefaultConfig() *Config
```

--------------------------------

### NewSwarm

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Initializes a new Swarm instance with an LLM client.

```APIDOC
## NewSwarm

### Description
Initializes a new Swarm instance with an LLM client.

### Signature
```go
func NewSwarm(apiKey string, provider llm.LLMProvider) *Swarm
```
```

--------------------------------

### Configure and Run Agents Concurrently

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Set up multiple agents with their configurations (messages, max turns, tool execution) and run them in parallel using ConcurrentSwarm. Includes context-based timeout and error handling.

```go
// Create a concurrent swarm
cs := swarmgo.NewConcurrentSwarm(apiKey)

// Configure multiple agents
configs := map[string]swarmgo.AgentConfig{
    "agent1": {
        Agent: agent1,
        Messages: []openai.ChatCompletionMessage{
            {Role: openai.ChatMessageRoleUser, Content: "Task 1"},
        },
        MaxTurns: 1,
        ExecuteTools: true,
    },
    "agent2": {
        Agent: agent2,
        Messages: []openai.ChatCompletionMessage{
            {Role: openai.ChatMessageRoleUser, Content: "Task 2"},
        },
        MaxTurns: 1,
        ExecuteTools: true,
    },
}

// Run agents concurrently with a timeout
ctx, cancel := context.WithTimeout(context.Background(), 5*time.Minute)
defer cancel()

results := cs.RunConcurrent(ctx, configs)

// Process results
for _, result := range results {
    if result.Error != nil {
        log.Printf("Error in %s: %v\n", result.AgentName, result.Error)
        continue
    }
    // Handle successful response
    fmt.Printf("Result from %s: %s\n", result.AgentName, result.Response)
}
```

--------------------------------

### NewSwarmWithHost

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Creates a Swarm with a custom host.

```APIDOC
## NewSwarmWithHost

### Description
Creates a Swarm with a custom host.

### Signature
```go
func NewSwarmWithHost(apiKey, host string, provider llm.LLMProvider) *Swarm
```
```

--------------------------------

### Define StreamHandler Interface

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Represents a handler for streaming responses. Defines methods for start, token, tool call, completion, and error events.

```go
type StreamHandler interface {
	OnStart()
	OnToken(token string)
	OnToolCall(toolCall llm.ToolCall)
	OnComplete(message llm.Message)
	OnError(err error)
}
```

--------------------------------

### Configure Agent with Dynamic Instructions

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Sets a function to dynamically generate instructions for an agent based on context variables.

```go
func (a *Agent) WithInstructionsFunc(f func(map[string]interface{}) string) *Agent
```

--------------------------------

### Build Workflow from Specification

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Constructs a concrete Workflow instance from a given WorkflowSpec.

```go
func (dwc *DynamicWorkflowCreator) BuildWorkflow(spec *WorkflowSpec) (*Workflow, error)
```

--------------------------------

### Get String Value from GraphState

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Retrieves a string value from the GraphState. Returns the string value and a second boolean indicating if the key was found and was of string type.

```go
func (s GraphState) GetString(key StateKey) (string, bool)
```

--------------------------------

### Get Boolean Value from GraphState

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Retrieves a boolean value from the GraphState. Returns the boolean value and a second boolean indicating if the key was found and was of boolean type.

```go
func (s GraphState) GetBool(key StateKey) (bool, bool)
```

--------------------------------

### Configure Agent with Functions

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Sets the list of functions that an agent can perform.

```go
func (a *Agent) WithFunctions(functions []AgentFunction) *Agent
```

--------------------------------

### Initialize New Swarm

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Initializes a new Swarm instance with an API key and an LLM provider.

```go
func NewSwarm(apiKey string, provider llm.LLMProvider) *Swarm
```

--------------------------------

### Initialize New Workflow

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Initializes a new Workflow instance with an API key and LLM provider. Added in v1.0.4.

```go
func NewWorkflow(apikey string, provider llm.LLMProvider, workflowType WorkflowType) *Workflow
```

--------------------------------

### Define StreamHandler Interface

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Implement the StreamHandler interface to process streaming responses. This interface includes methods for handling stream start, tokens, tool calls, completion, and errors.

```go
type StreamHandler interface {
    OnStart()
    OnToken(token string)
    OnToolCall(toolCall openai.ToolCall)
    OnComplete(message openai.ChatCompletionMessage)
    OnError(err error)
}
```

--------------------------------

### NewOpenAILLMWithHost

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

NewOpenAILLMWithHost creates a new OpenAI LLM client with a specified host. It requires an API key.

```APIDOC
## NewOpenAILLMWithHost

### Description
Creates a new OpenAI LLM client with a specified host.

### Parameters
#### Path Parameters
- **apiKey** (string) - Required - The API key for authenticating with the OpenAI service.
- **host** (string) - Required - The host for the OpenAI service.

### Returns
- *OpenAILLM - A pointer to the initialized OpenAILLM client.
```

--------------------------------

### OpenAILLM Client Creation with Custom Host

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new OpenAI LLM client with a specified host, useful for custom endpoints or proxies.

```go
func NewOpenAILLMWithHost(apiKey string, host string) *OpenAILLM
```

--------------------------------

### Create and Execute Workflow

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

A convenience method to create and execute a workflow based on user tasks in a single step.

```go
func (dwc *DynamicWorkflowCreator) CreateAndExecuteWorkflow(ctx context.Context, userTask string) (*WorkflowResult, error)
```

--------------------------------

### Create New Agent

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Creates a new agent instance with an initialized memory store.

```go
func NewAgent(name, model string, provider llm.LLMProvider) *Agent
```

--------------------------------

### NewOpenAILLM

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new OpenAI LLM client. Requires an API key for authentication.

```APIDOC
## NewOpenAILLM

### Description
Creates a new OpenAI LLM client. This function initializes the client with the provided API key.

### Method Signature
```go
func NewOpenAILLM(apiKey string) *OpenAILLM
```

### Parameters
- `apiKey` (string): The API key for authenticating with the OpenAI service.
```

--------------------------------

### OpenAILLM Client Creation

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new OpenAI LLM client. Requires an API key for authentication.

```go
func NewOpenAILLM(apiKey string) *OpenAILLM
```

--------------------------------

### RunDemoLoopWithConfig

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Initiates an interactive CLI loop with custom configuration options for agent interaction.

```APIDOC
## RunDemoLoopWithConfig

### Description
RunDemoLoopWithConfig starts an interactive CLI loop with custom configuration.

### Signature
```go
func RunDemoLoopWithConfig(client *Swarm, agent *Agent, config *DemoLoopConfig)
```
```

--------------------------------

### Check Swarm Initialization

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Returns true if the Swarm instance is properly initialized, false otherwise.

```go
func (s *Swarm) IsInitialized() bool
```

--------------------------------

### NewOpenAILLMWithHost

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new OpenAI LLM client with a specified host. Allows overriding the default OpenAI API endpoint.

```APIDOC
## NewOpenAILLMWithHost

### Description
Creates a new OpenAI LLM client, allowing the specification of a custom host (API endpoint).

### Method Signature
```go
func NewOpenAILLMWithHost(apiKey string, host string) *OpenAILLM
```

### Parameters
- `apiKey` (string): The API key for authenticating with the OpenAI service.
- `host` (string): The custom host or API endpoint to use for the OpenAI service.
```

--------------------------------

### Swarm Initialization

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Provides functions to create new Swarm instances with different configuration options.

```APIDOC
## Swarm Initialization

### Description
Functions to create and initialize a Swarm instance.

### Functions

*   **NewSwarm**: Creates a new Swarm instance with an API key and LLM provider.
    ```go
    func NewSwarm(apiKey string, provider llm.LLMProvider) *Swarm
    ```
*   **NewSwarmWithConfig**: Creates a new Swarm instance with an API key, LLM provider, and custom configuration.
    ```go
    func NewSwarmWithConfig(apiKey string, provider llm.LLMProvider, config *Config) *Swarm
    ```
*   **NewSwarmWithHost**: Creates a new Swarm instance specifying the API key, host, and LLM provider.
    ```go
    func NewSwarmWithHost(apiKey, host string, provider llm.LLMProvider) *Swarm
    ```
```

--------------------------------

### DeepSeekLLM Client Creation

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new DeepSeek LLM client. Requires an API key for authentication.

```go
func NewDeepSeekLLM(apiKey string) *DeepSeekLLM
```

--------------------------------

### NewGeminiLLM

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new Gemini LLM client. Requires an API key and accepts optional configuration options.

```APIDOC
## NewGeminiLLM

### Description
Creates a new Gemini LLM client. This function initializes the client with the provided API key and optional configuration settings.

### Method Signature
```go
func NewGeminiLLM(apiKey string, opts ...GeminiOptions) (*GeminiLLM, error)
```

### Parameters
- `apiKey` (string): The API key for authenticating with the Gemini service.
- `opts` (...GeminiOptions): Optional configuration options for the Gemini model.
```

--------------------------------

### OllamaLLM Client Creation

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new Ollama LLM client. For use with a default local Ollama instance.

```go
func NewOllamaLLM() (*OllamaLLM, error)
```

--------------------------------

### OllamaLLM Client Creation with Custom URL

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new Ollama LLM client with a custom base URL, allowing connection to a remote Ollama instance.

```go
func NewOllamaLLMWithURL(baseURL string) (*OllamaLLM, error)
```

--------------------------------

### Generate Workflow Specification from Task

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Generates a workflow specification based on a user's task description.

```go
func (dwc *DynamicWorkflowCreator) CreateWorkflowFromTask(ctx context.Context, userTask string) (*WorkflowSpec, error)
```

--------------------------------

### Manage Agent Memory in SwarmGo

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Demonstrates creating an agent with memory, explicitly storing memories with context and importance, and retrieving/searching memories.

```go
import "time"

// Create a new agent with memory capabilities
agent := swarmgo.NewAgent("MyAgent", "gpt-4")

// Memory is automatically managed for conversations and tool calls
// You can also explicitly store memories:
memory := swarmgo.Memory{
    Content:    "User prefers dark mode",
    Type:       "preference",
    Context:    map[string]interface{}{"setting": "ui"},
    Timestamp:  time.Now(),
    Importance: 0.8,
}
agent.Memory.AddMemory(memory)

// Retrieve recent memories
recentMemories := agent.Memory.GetRecentMemories(5)

// Search specific types of memories
preferences := agent.Memory.SearchMemories("preference", nil)

```

--------------------------------

### Workflow Initialization

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Functions to create new Workflow instances.

```APIDOC
## Workflow Initialization

### Description
Functions to create and initialize a Workflow instance.

### Functions

*   **NewWorkflow**: Creates a new Workflow instance with an API key, LLM provider, and workflow type.
    ```go
    func NewWorkflow(apikey string, provider llm.LLMProvider, workflowType WorkflowType) *Workflow
    ```
```

--------------------------------

### Create New Claude LLM Client

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new Claude LLM client. Requires an API key for authentication.

```go
func NewClaudeLLM(apiKey string) *ClaudeLLM
```

--------------------------------

### OllamaLLM CreateChatCompletion

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Implements the LLM interface for Ollama chat completions.

```go
func (o *OllamaLLM) CreateChatCompletion(ctx context.Context, req ChatCompletionRequest) (ChatCompletionResponse, error)
```

--------------------------------

### Run Agent Execution

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

The main entry point for agent execution. Manages the workflow with options for context, messages, variables, model override, streaming, debugging, max turns, and tool execution.

```go
func (s *Swarm) Run(
	ctx context.Context,
	agent *Agent,
	messages []llm.Message,
	contextVariables map[string]interface{},
	modelOverride string,
	stream bool,
	debug bool,
	maxTurns int,
	executeTools bool,
) (Response, error)
```

--------------------------------

### NewGeminiLLM

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

NewGeminiLLM creates a new Gemini LLM client. It requires an API key and optional configuration options.

```APIDOC
## NewGeminiLLM

### Description
Creates a new Gemini LLM client.

### Parameters
#### Path Parameters
- **apiKey** (string) - Required - The API key for authenticating with the Gemini service.
- **opts** (...GeminiOptions) - Optional - Additional configuration options for the Gemini client.

### Returns
- **(*GeminiLLM, error)** - A pointer to the initialized GeminiLLM client and an error if initialization fails.
```

--------------------------------

### GeminiLLM Client Creation

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new Gemini LLM client. Accepts an API key and optional configuration options.

```go
func NewGeminiLLM(apiKey string, opts ...GeminiOptions) (*GeminiLLM, error)
```

--------------------------------

### NewDeepSeekLLM

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new DeepSeek LLM client. Requires an API key for authentication.

```APIDOC
## NewDeepSeekLLM

### Description
Creates a new DeepSeek LLM client. This function initializes the client with the provided API key.

### Method Signature
```go
func NewDeepSeekLLM(apiKey string) *DeepSeekLLM
```

### Parameters
- `apiKey` (string): The API key for authenticating with the DeepSeek service.
```

--------------------------------

### GeminiLLM CreateChatCompletion

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Implements the LLM interface for Gemini chat completions.

```go
func (g *GeminiLLM) CreateChatCompletion(ctx context.Context, req ChatCompletionRequest) (ChatCompletionResponse, error)
```

--------------------------------

### Create Human Input Node Function

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Creates a node that collects input from a human user. Requires graph, ID, and a prompt string.

```go
func CreateHumanInputNode(g *Graph, id NodeID, prompt string) *Node
```

--------------------------------

### Configure LLM Client

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

ClientConfig holds settings for connecting to a Language Model (LLM) provider. It includes authentication, base URL, and model mapping.

```go
type ClientConfig struct {
	Provider           llm.LLMProvider
	AuthToken          string
	BaseURL            string
	OrgID              string
	APIVersion         string
	AssistantVersion   string
	ModelMapperFunc    func(model string) string // replace model to provider-specific deployment name
	HTTPClient         *http.Client
	EmptyMessagesLimit uint
	Options            map[string]interface{} // Additional provider-specific options
}
```

--------------------------------

### NewWorkflow

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Initializes a new Workflow instance with the provided API key, LLM provider, and workflow type. Added in v1.0.4.

```APIDOC
## NewWorkflow

### Description
NewWorkflow initializes a new Workflow instance.

### Method
func NewWorkflow(apikey string, provider llm.LLMProvider, workflowType WorkflowType) *Workflow
```

--------------------------------

### IsInitialized

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Returns whether the Swarm is properly initialized.

```APIDOC
## IsInitialized

### Description
Returns whether the Swarm is properly initialized.

### Signature
```go
func (s *Swarm) IsInitialized() bool
```
```

--------------------------------

### OllamaLLM CreateChatCompletionStream

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Implements the LLM interface for Ollama streaming.

```go
func (o *OllamaLLM) CreateChatCompletionStream(ctx context.Context, req ChatCompletionRequest) (ChatCompletionStream, error)
```

--------------------------------

### Use Streaming Response with Custom Handler

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Initiate a streaming response from a client by providing an agent, messages, and a custom StreamHandler. Ensure the API key and model are correctly configured.

```go
client := swarmgo.NewSwarm("YOUR_OPENAI_API_KEY", llm.OpenAI)

agent := &swarmgo.Agent{
    Name:         "FileAnalyzer",
    Instructions: "You are an assistant that analyzes files.",
    Model:        "gpt-4",
}

handler := &CustomStreamHandler{}
er := client.StreamingResponse(
    context.Background(),
    agent,
    messages,
    nil,
    "",
    handler,
    true,
)
```

--------------------------------

### NewAgent

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Creates a new agent with a specified name, model, and LLM provider, initializing its memory store.

```APIDOC
## NewAgent

### Description
Creates a new agent with initialized memory store.

### Signature
```go
func NewAgent(name, model string, provider llm.LLMProvider) *Agent
```
```

--------------------------------

### OllamaLLM.CreateChatCompletion

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a chat completion with the Ollama LLM. Implements the LLM interface for Ollama.

```APIDOC
## OllamaLLM.CreateChatCompletion

### Description
Creates a chat completion using the Ollama LLM. This method adheres to the `LLM` interface for Ollama.

### Method Signature
```go
func (o *OllamaLLM) CreateChatCompletion(ctx context.Context, req ChatCompletionRequest) (ChatCompletionResponse, error)
```

### Parameters
- `ctx` (context.Context): The context for the request.
- `req` (ChatCompletionRequest): The request object containing chat completion parameters.
```

--------------------------------

### Implement Agent Handoff Function

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Create a function to transfer conversations to another agent. This function should return a swarmgo.Result with the target agent.

```go
func transferToAnotherAgent(args map[string]interface{}, contextVariables map[string]interface{}) swarmgo.Result {
	anotherAgent := &swarmgo.Agent{
		Name:         "AnotherAgent",
		Instructions: "You are another agent.",
		Model:        "gpt-3.5-turbo",
	}
	return swarmgo.Result{
		Agent: anotherAgent,
		Value: "Transferring to AnotherAgent.",
	}
}
```

--------------------------------

### Define a Tool Function for an Agent

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Illustrates how to define a Go function that can be used as a tool by an agent. The function accepts arguments and context variables, and returns a SwarmGo Result.

```go
func getWeather(args map[string]interface{}, contextVariables map[string]interface{}) swarmgo.Result {
	location := args["location"].(string)
	// Simulate fetching weather data
	return swarmgo.Result{
		Value: fmt.Sprintf(`{"temp": 67, "unit": "F", "location": "%s"}`, location),
	}
}

```

--------------------------------

### Create New DynamicWorkflowCreator

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Initializes a new DynamicWorkflowCreator with an API key and an LLM provider.

```go
func NewDynamicWorkflowCreator(apiKey string, provider llm.LLMProvider) *DynamicWorkflowCreator
```

--------------------------------

### Create Concurrent Swarm Instance

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

NewConcurrentSwarm initializes a new ConcurrentSwarm with the provided API key and LLM provider.

```go
func NewConcurrentSwarm(apiKey string, provider llm.LLMProvider) *ConcurrentSwarm
```

--------------------------------

### DeepSeekLLM CreateChatCompletion

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Implements the LLM interface for DeepSeek chat completions.

```go
func (l *DeepSeekLLM) CreateChatCompletion(ctx context.Context, req ChatCompletionRequest) (ChatCompletionResponse, error)
```

--------------------------------

### Demo Loop Configuration

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

DemoLoopConfig holds settings for controlling the behavior of a demo loop, such as timeouts, history management, and output formatting.

```go
type DemoLoopConfig struct {
	Timeout             time.Duration // Timeout for each agent execution
	MaxHistoryMessages  int           // Maximum number of messages to keep in history
	MaxInputLength      int           // Maximum length of user input
	ShowFunctionResults bool          // Whether to display function results
	ColorOutput         bool          // Whether to use color in output
	Debug               bool          // Whether to show debug information
	SaveHistory         bool          // Whether to save conversation history to file
	HistoryFile         string        // Path to file for saving history
}
```

--------------------------------

### Create New MemoryStore

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Factory function to create a new MemoryStore instance. Allows specifying the maximum capacity for short-term memory.

```go
func NewMemoryStore(maxShortTerm int) *MemoryStore
```

--------------------------------

### NewOpenAILLM

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

NewOpenAILLM creates a new OpenAI LLM client. It requires an API key for authentication.

```APIDOC
## NewOpenAILLM

### Description
Creates a new OpenAI LLM client.

### Parameters
#### Path Parameters
- **apiKey** (string) - Required - The API key for authenticating with the OpenAI service.

### Returns
- *OpenAILLM - A pointer to the initialized OpenAILLM client.
```

--------------------------------

### DefaultDemoLoopConfig

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Returns default configuration for the demo loop.

```APIDOC
## DefaultDemoLoopConfig

### Description
Returns default configuration for demo loop.

### Signature
```go
func DefaultDemoLoopConfig() *DemoLoopConfig
```
```

--------------------------------

### GeminiLLM CreateChatCompletionStream

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Implements the LLM interface for Gemini streaming.

```go
func (g *GeminiLLM) CreateChatCompletionStream(ctx context.Context, req ChatCompletionRequest) (ChatCompletionStream, error)
```

--------------------------------

### GraphRunner Methods

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Methods for creating, registering, and executing graphs.

```APIDOC
## NewGraphRunner

### Description
Creates a new graph runner.

### Signature
```go
func NewGraphRunner() *GraphRunner
```
```

```APIDOC
## GraphRunner.ExecuteGraph

### Description
Runs a graph with the given initial state.

### Signature
```go
func (r *GraphRunner) ExecuteGraph(ctx context.Context, graphID string, initialState GraphState) (GraphState, error)
```
```

```APIDOC
## GraphRunner.RegisterGraph

### Description
Adds a graph to the runner.

### Signature
```go
func (r *GraphRunner) RegisterGraph(graph *Graph)
```
```

--------------------------------

### LLMProvider Constants

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Defines constants for different LLM providers.

```go
type LLMProvider string

const (
	OpenAI          LLMProvider = "OPEN_AI"
	Azure           LLMProvider = "AZURE"
	AzureAD         LLMProvider = "AZURE_AD"
	CloudflareAzure LLMProvider = "CLOUDFLARE_AZURE"
	Gemini          LLMProvider = "GEMINI"
	Claude          LLMProvider = "CLAUDE"
	Ollama          LLMProvider = "OLLAMA"
	DeepSeek        LLMProvider = "DEEPSEEK"
)
```

--------------------------------

### Set Graph Entry Point with GraphBuilder

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Sets the designated entry point for the workflow graph. Only one entry point is typically allowed.

```go
func (b *GraphBuilder) WithEntryPoint(nodeID NodeID) *GraphBuilder
```

--------------------------------

### Create Agent Node Function

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Helper function to create common agent node types. Requires graph, ID, name, instructions, model, functions, and LLM provider.

```go
func CreateAgentNode(g *Graph, id NodeID, name string, instructions string, model string, functions []AgentFunction, provider llm.LLMProvider) *Node
```

--------------------------------

### Run Agent with Messages

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Executes an agent with a given set of messages and retrieves the response. This method handles the interaction with the LLM based on the agent's configuration.

```go
messages := []openai.ChatCompletionMessage{
	{Role: "user", Content: "Hello!"},
}

ctx := context.Background()
response, err := client.Run(ctx, agent, messages, nil, "", false, false, 5, true)
if err != nil {
	log.Fatalf("Error: %v", err)
}

fmt.Println(response.Messages[len(response.Messages)-1].Content)

```

--------------------------------

### Set Cycle Handling Strategy

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Configures how the workflow should handle detected cycles. Added in v1.0.4.

```go
func (wf *Workflow) SetCycleHandling(handling CycleHandling)
```

--------------------------------

### Agent.WithConfig

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Sets the configuration for an agent, allowing customization of its behavior and parameters.

```APIDOC
## Agent.WithConfig

### Description
Sets the configuration for the agent.

### Signature
```go
func (a *Agent) WithConfig(config *ClientConfig) *Agent
```
```

--------------------------------

### Create New GraphRunner

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Factory function to instantiate a new GraphRunner. Initializes the runner for graph execution.

```go
func NewGraphRunner() *GraphRunner
```

--------------------------------

### NewOllamaLLM

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a new Ollama LLM client. This function initializes the client with default settings.

```APIDOC
## NewOllamaLLM

### Description
Creates a new Ollama LLM client using default configuration.

### Method Signature
```go
func NewOllamaLLM() (*OllamaLLM, error)
```

### Returns
- `*OllamaLLM`: A pointer to the initialized OllamaLLM client.
- `error`: An error if the client creation fails.
```

--------------------------------

### MemoryStore Methods

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Methods for managing agent memories, including adding, retrieving, searching, and persisting memories.

```APIDOC
## NewMemoryStore

### Description
Creates a new memory store with default settings.

### Signature
```go
func NewMemoryStore(maxShortTerm int) *MemoryStore
```
```

```APIDOC
## MemoryStore.AddMemory

### Description
Adds a new memory to both short and long-term storage.

### Signature
```go
func (ms *MemoryStore) AddMemory(memory Memory)
```
```

```APIDOC
## MemoryStore.GetRecentMemories

### Description
Retrieves the n most recent memories.

### Signature
```go
func (ms *MemoryStore) GetRecentMemories(n int) []Memory
```
```

```APIDOC
## MemoryStore.LoadMemories

### Description
Loads memories from JSON data.

### Signature
```go
func (ms *MemoryStore) LoadMemories(data []byte) error
```
```

```APIDOC
## MemoryStore.SearchMemories

### Description
Searches for memories based on type and context.

### Signature
```go
func (ms *MemoryStore) SearchMemories(memoryType string, context map[string]interface{}) []Memory
```
```

```APIDOC
## MemoryStore.SerializeMemories

### Description
Serializes all memories to JSON.

### Signature
```go
func (ms *MemoryStore) SerializeMemories() ([]byte, error)
```
```

--------------------------------

### DeepSeekLLM CreateChatCompletionStream

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Implements the LLM interface for DeepSeek streaming.

```go
func (l *DeepSeekLLM) CreateChatCompletionStream(ctx context.Context, req ChatCompletionRequest) (ChatCompletionStream, error)
```

--------------------------------

### GeminiOptions Structure

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Contains configuration options for the Gemini model, including model name, harm thresholds, and safety settings.

```go
type GeminiOptions struct {
	Model          string
	HarmThreshold  genai.HarmBlockThreshold
	SafetySettings []*genai.SafetySetting
}
```

--------------------------------

### Swarm Configuration Options

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Config holds various settings for the Swarm, including retry logic, timeouts, token limits, and failure handling strategies.

```go
type Config struct {
	MaxRetries        int
	RetryBackoff      time.Duration
	RequestTimeout    time.Duration
	MaxTokens         int
	DefaultModel      string
	Debug             bool
	LogLevel          LogLevel
	TokenLimits       map[string]int // Model-specific token limits
	FailureHandlers   []FailureHandler
	RateLimitStrategy RateLimitStrategy
}
```

--------------------------------

### OpenAILLM.CreateChatCompletion

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

CreateChatCompletion implements the LLM interface for OpenAI. It sends a chat completion request and returns the response.

```APIDOC
## OpenAILLM.CreateChatCompletion

### Description
Implements the LLM interface for OpenAI to create a chat completion.

### Method
- OpenAILLM.CreateChatCompletion

### Parameters
#### Path Parameters
- **ctx** (context.Context) - Required - The context for the request.
- **req** (ChatCompletionRequest) - Required - The chat completion request object.

### Returns
- **ChatCompletionResponse** - The response from the chat completion.
- **error** - An error if the request fails.
```

--------------------------------

### GeminiLLM.CreateChatCompletion

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a chat completion with the Gemini LLM. Implements the LLM interface for Gemini.

```APIDOC
## GeminiLLM.CreateChatCompletion

### Description
Creates a chat completion using the Gemini LLM. This method adheres to the `LLM` interface for Gemini.

### Method Signature
```go
func (g *GeminiLLM) CreateChatCompletion(ctx context.Context, req ChatCompletionRequest) (ChatCompletionResponse, error)
```

### Parameters
- `ctx` (context.Context): The context for the request.
- `req` (ChatCompletionRequest): The request object containing chat completion parameters.
```

--------------------------------

### ClaudeLLM CreateChatCompletionStream

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Implements the LLM interface for Claude streaming.

```go
func (c *ClaudeLLM) CreateChatCompletionStream(ctx context.Context, req ChatCompletionRequest) (ChatCompletionStream, error)
```

--------------------------------

### Add a Function (Tool) to an Agent

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Configures an agent to use a specific function by providing its name, description, parameters schema, and the function implementation. This allows the agent to call the defined function when needed.

```go
agent.Functions = []swarmgo.AgentFunction{
	{
		Name:        "getWeather",
		Description: "Get the current weather in a given location.",
		Parameters: map[string]interface{}{
			"type": "object",
			"properties": map[string]interface{}{
				"location": map[string]interface{}{
					"type":        "string",
					"description": "The city to get the weather for.",
				},
			},
			"required": []interface{}{"location"},
		},
		Function: getWeather,
	},
}

```

--------------------------------

### DeepSeekLLM.CreateChatCompletion

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a chat completion with the DeepSeek LLM. Implements the LLM interface for DeepSeek.

```APIDOC
## DeepSeekLLM.CreateChatCompletion

### Description
Creates a chat completion using the DeepSeek LLM. This method adheres to the `LLM` interface for DeepSeek.

### Method Signature
```go
func (l *DeepSeekLLM) CreateChatCompletion(ctx context.Context, req ChatCompletionRequest) (ChatCompletionResponse, error)
```

### Parameters
- `ctx` (context.Context): The context for the request.
- `req` (ChatCompletionRequest): The request object containing chat completion parameters.
```

--------------------------------

### OllamaLLM.CreateChatCompletionStream

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

Creates a chat completion stream with the Ollama LLM. Implements the LLM interface for Ollama streaming.

```APIDOC
## OllamaLLM.CreateChatCompletionStream

### Description
Creates a chat completion stream with the Ollama LLM. This method implements the `LLM` interface for handling streaming responses from Ollama.

### Method Signature
```go
func (o *OllamaLLM) CreateChatCompletionStream(ctx context.Context, req ChatCompletionRequest) (ChatCompletionStream, error)
```

### Parameters
- `ctx` (context.Context): The context for the request.
- `req` (ChatCompletionRequest): The request object containing chat completion parameters.
```

--------------------------------

### Create Router Node Function

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Creates a node that routes to different destinations based on content. Requires graph, ID, and a map of destination names to NodeIDs.

```go
func CreateRouterNode(g *Graph, id NodeID, destinations map[string]NodeID) *Node
```

--------------------------------

### Agent Struct Definition

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Defines the Agent structure, including its name, model, provider, configuration, instructions, functions, memory, and tool call settings.

```go
type Agent struct {
	Name              string                                               // The name of the agent.
	Model             string                                               // The model identifier.
	Provider          llm.LLMProvider                                      // The LLM provider to use.
	Config            *ClientConfig                                        // Provider-specific configuration.
	Instructions      string                                               // Static instructions for the agent.
	InstructionsFunc  func(contextVariables map[string]interface{}) string // Function to generate dynamic instructions based on context.
	Functions         []AgentFunction                                      // A list of functions the agent can perform.
	Memory            *MemoryStore                                         // Memory store for the agent.
	ParallelToolCalls bool                                                 // Whether to allow parallel tool calls.
}
```

--------------------------------

### Workflow Structure

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Represents a collection of agents and their connections. Added in v1.0.4.

```go
type Workflow struct {
	// contains filtered or unexported fields
}
```

--------------------------------

### Run Agents Concurrently in Order

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

RunConcurrentOrdered executes agents concurrently but returns results in the order specified by the input configuration slice.

```go
func (cs *ConcurrentSwarm) RunConcurrentOrdered(ctx context.Context, orderedConfigs []struct {
	Name   string
	Config AgentConfig
}) []ConcurrentResult
```

--------------------------------

### OllamaLLM.CreateChatCompletionStream

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo/llm

CreateChatCompletionStream implements the LLM interface for Ollama for streaming chat completions.

```APIDOC
## OllamaLLM.CreateChatCompletionStream

### Description
Implements the LLM interface for Ollama to create a streaming chat completion.

### Method
- OllamaLLM.CreateChatCompletionStream

### Parameters
#### Path Parameters
- **ctx** (context.Context) - Required - The context for the request.
- **req** (ChatCompletionRequest) - Required - The chat completion request object.

### Returns
- **ChatCompletionStream** - An interface for receiving streaming chat completion responses.
- **error** - An error if the request fails.
```

--------------------------------

### Build Graph from Builder

Source: https://pkg.go.dev/github.com/prathyushnallamothu/swarmgo

Finalizes the graph construction and returns the built Graph object.

```go
func (b *GraphBuilder) Build() *Graph
```
Tokens:
Raw