agent

package
v0.9.1 Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Apr 23, 2026 License: AGPL-3.0 Imports: 27 Imported by: 0

Documentation

Index

Constants

This section is empty.

Variables

This section is empty.

Functions

func QuickStartOpenAI 

func QuickStartOpenAI(privateKey, openaiKey string) error

QuickStartOpenAI creates and runs an OpenAI agent with minimal configuration This is the absolute simplest way to start - just provide your keys!

Example: 

agent.QuickStartOpenAI("0xYourPrivateKey", "sk-YourOpenAIKey")

func SubmitForReview added in v0.6.5 

func SubmitForReview(backendURL, agentID, creatorWallet string, tokenID uint64, headers ...map[string]string) error

func WithdrawPublic added in v0.6.5 

func WithdrawPublic(backendURL, agentID, creatorWallet string, tokenID uint64) error

WithdrawPublic withdraws a public agent back to private visibility on the Teneo network.

Only agents with review status "public" can be withdrawn. After withdrawal the agent returns to "private" status and must be re-submitted for review to become public again.

This is a standalone utility that can be called from any context — no running agent required.

Parameters:

  • backendURL: The Teneo backend URL (e.g. "https://backend.developer.chatroom.teneo-protocol.ai")
  • agentID: The agent's ID from your metadata JSON (agentId field)
  • creatorWallet: The Ethereum wallet address that owns the agent's NFT
  • tokenID: The NFT token ID for on-chain ownership verification

Example: 

err := agent.WithdrawPublic(
    "https://backend.developer.chatroom.teneo-protocol.ai",
    "my-agent-id",
    "0xYourWalletAddress",
    42,
)

HTTP API equivalent (for non-Go clients): 

POST {backendURL}/api/agents/{agent-id}/withdraw-public
Content-Type: application/json

{
    "creator_wallet": "0xYourWalletAddress",
    "token_id": 42
}

Types

type Agent 

type Agent struct {
	// contains filtered or unexported fields
}

Agent represents a Teneo agent instance

func NewAgent 

func NewAgent(config *Config, handler types.AgentHandler) (*Agent, error)

NewAgent creates a new Teneo agent instance

func (*Agent) Close 

func (a *Agent) Close() error

Close releases resources used by the agent

func (*Agent) GetAddress 

func (a *Agent) GetAddress() string

GetAddress returns the agent's Ethereum address

func (*Agent) GetAuthToken 

func (a *Agent) GetAuthToken() (string, error)

GetAuthToken generates an authentication token for the agent

func (*Agent) GetBusinessCard 

func (a *Agent) GetBusinessCard() (*types.BusinessCard, error)

GetBusinessCard returns the agent's business card information

func (*Agent) GetCapabilities 

func (a *Agent) GetCapabilities() []string

GetCapabilities returns the agent's capabilities

func (*Agent) IsRunning 

func (a *Agent) IsRunning() bool

IsRunning returns whether the agent is currently running

func (*Agent) Run 

func (a *Agent) Run(ctx context.Context) error

Run starts the agent and blocks until it's stopped

func (*Agent) SetActive 

func (a *Agent) SetActive(active bool) error

SetActive sets the agent's active status

func (*Agent) Shutdown 

func (a *Agent) Shutdown()

Shutdown gracefully shuts down the agent

func (*Agent) UpdateMetadata 

func (a *Agent) UpdateMetadata(description, contactInfo, pricingModel, version string) error

UpdateMetadata updates the agent's metadata on the blockchain

type Config 

type Config struct {
	// Basic agent info
	AgentID      string   `json:"agent_id"`
	Name         string   `json:"name"`
	Description  string   `json:"description"`
	Image        string   `json:"image"`
	Version      string   `json:"version"`
	Capabilities []string `json:"capabilities"`
	ContactInfo  string   `json:"contact_info"`
	PricingModel string   `json:"pricing_model"`

	// CapabilityDetails provides full capability objects with descriptions.
	// When set, these take precedence over the string-only Capabilities field
	// during deployment and metadata serialization. If empty, Capabilities
	// strings are used with empty descriptions for backward compatibility.
	CapabilityDetails []types.Capability `json:"capability_details,omitempty"`

	// Profile metadata (optional, sent to backend during deploy/update)
	ShortDescription string `json:"short_description,omitempty"` // Brief one-line summary
	TutorialURL      string `json:"tutorial_url,omitempty"`      // YouTube/video tutorial URL

	// Interface configuration
	InterfaceType  string `json:"interface_type"`
	ResponseFormat string `json:"response_format"`

	// Network configuration
	WebSocketURL     string        `json:"websocket_url"`
	ReconnectEnabled bool          `json:"reconnect_enabled"`
	ReconnectDelay   time.Duration `json:"reconnect_delay"`
	MaxReconnects    int           `json:"max_reconnects"`
	MessageTimeout   time.Duration `json:"message_timeout"`
	PingInterval     time.Duration `json:"ping_interval"`
	HandshakeTimeout time.Duration `json:"handshake_timeout"`

	// Health monitoring
	HealthEnabled bool `json:"health_enabled"`
	HealthPort    int  `json:"health_port"`

	// Authentication
	PrivateKey   string `json:"private_key"`
	OwnerAddress string `json:"owner_address"`
	NFTTokenID   string `json:"nft_token_id"`

	// Room configuration
	Room string `json:"room"`

	// Blockchain configuration
	EthereumRPC        string `json:"ethereum_rpc"`
	NFTContractAddress string `json:"nft_contract_address"`

	// Task processing
	MaxConcurrentTasks int `json:"max_concurrent_tasks"`
	TaskTimeout        int `json:"task_timeout"`
	TaskCheckInterval  int `json:"task_check_interval"`

	// Rate limiting
	RateLimitPerMinute int `json:"rate_limit_per_minute"` // 0 = unlimited

	// Redis cache configuration
	RedisEnabled   bool   `json:"redis_enabled"`    // Enable Redis caching
	RedisAddress   string `json:"redis_address"`    // Redis server address (e.g., "localhost:6379")
	RedisUsername  string `json:"redis_username"`   // Redis ACL username (Redis 6+, empty for legacy auth)
	RedisPassword  string `json:"redis_password"`   // Redis password (empty if no password)
	RedisDB        int    `json:"redis_db"`         // Redis database number (0-15)
	RedisKeyPrefix string `json:"redis_key_prefix"` // Prefix for all cache keys
	RedisUseTLS    bool   `json:"redis_use_tls"`    // Enable TLS/SSL (required for managed Redis)

	// Slack alerting configuration
	SlackWebhookURL           string `json:"slack_webhook_url"`            // Slack webhook URL for failure alerts
	SlackAlertThrottleSeconds int    `json:"slack_alert_throttle_seconds"` // Throttle window in seconds (default: 60)

	// Additional HTTP headers sent during WebSocket handshake
	AdditionalHeaders map[string]string `json:"additional_headers,omitempty"`
}

Config represents the configuration for a Teneo agent

func DefaultConfig 

func DefaultConfig() *Config

DefaultConfig returns a default configuration

func (*Config) LoadFromEnv 

func (c *Config) LoadFromEnv() error

LoadFromEnv loads configuration from environment variables

func (*Config) ResolveCapabilities added in v0.6.8 

func (c *Config) ResolveCapabilities() []types.Capability

ResolveCapabilities returns the full Capability objects for this config. If CapabilityDetails is set, it takes precedence. Otherwise, Capabilities strings are converted to Capability objects with empty descriptions.

func (*Config) Validate 

func (c *Config) Validate() error

Validate validates the configuration

type EnhancedAgent 

type EnhancedAgent struct {
	// contains filtered or unexported fields
}

EnhancedAgent represents a fully functional Teneo network agent with all capabilities

func NewEnhancedAgent 

func NewEnhancedAgent(config *EnhancedAgentConfig) (*EnhancedAgent, error)

NewEnhancedAgent creates a new enhanced agent with network capabilities

func NewSimpleOpenAIAgent 

func NewSimpleOpenAIAgent(config *SimpleOpenAIAgentConfig) (*EnhancedAgent, error)

NewSimpleOpenAIAgent creates a fully configured Teneo agent powered by OpenAI in just a few lines This is the easiest way to get started with the Teneo Agent SDK.

Example: 

agent, err := agent.NewSimpleOpenAIAgent(&agent.SimpleOpenAIAgentConfig{
    PrivateKey: "0x...",
    OpenAIKey:  "sk-...",
})
if err != nil {
    log.Fatal(err)
}
agent.Run()

func (*EnhancedAgent) GetActiveTaskCount 

func (a *EnhancedAgent) GetActiveTaskCount() int

GetActiveTaskCount implements the health.StatusGetter interface

func (*EnhancedAgent) GetAuthManager 

func (a *EnhancedAgent) GetAuthManager() *auth.Manager

GetAuthManager returns the auth manager

func (*EnhancedAgent) GetCache 

func (a *EnhancedAgent) GetCache() cache.AgentCache

GetCache returns the agent cache instance This allows agent implementations to access the cache for persistent storage

func (*EnhancedAgent) GetConfig 

func (a *EnhancedAgent) GetConfig() *Config

GetConfig returns the agent configuration

func (*EnhancedAgent) GetNetworkClient 

func (a *EnhancedAgent) GetNetworkClient() *network.NetworkClient

GetNetworkClient returns the network client

func (*EnhancedAgent) GetTaskCoordinator 

func (a *EnhancedAgent) GetTaskCoordinator() *network.TaskCoordinator

GetTaskCoordinator returns the task coordinator

func (*EnhancedAgent) GetUptime 

func (a *EnhancedAgent) GetUptime() time.Duration

GetUptime implements the health.StatusGetter interface

func (*EnhancedAgent) IsAuthenticated 

func (a *EnhancedAgent) IsAuthenticated() bool

IsAuthenticated implements the health.StatusGetter interface

func (*EnhancedAgent) IsConnected 

func (a *EnhancedAgent) IsConnected() bool

IsConnected implements the health.StatusGetter interface

func (*EnhancedAgent) IsRunning 

func (a *EnhancedAgent) IsRunning() bool

IsRunning returns whether the agent is currently running

func (*EnhancedAgent) Run 

func (a *EnhancedAgent) Run() error

Run runs the agent until interrupted

func (*EnhancedAgent) Start 

func (a *EnhancedAgent) Start() error

Start starts the enhanced agent with all its components

func (*EnhancedAgent) Stop 

func (a *EnhancedAgent) Stop() error

Stop gracefully stops the enhanced agent

func (*EnhancedAgent) SubmitForReview added in v0.6.5 

func (a *EnhancedAgent) SubmitForReview() error

SubmitForReview submits the agent for public visibility review on the Teneo network. The agent must have been deployed, connected at least once, and be currently online. Review can take up to 72 hours. The agent must stay online during review.

func (*EnhancedAgent) SubmitForReviewDetailed added in v0.8.2 

func (a *EnhancedAgent) SubmitForReviewDetailed() (*SubmitForReviewResult, error)

SubmitForReview submits the agent for public visibility review on the Teneo network. The agent must have been deployed, connected at least once, and be currently online. Review can take up to 72 hours. The agent must stay online during review.

func (*EnhancedAgent) UpdateCapabilities 

func (a *EnhancedAgent) UpdateCapabilities(capabilities []string)

UpdateCapabilities updates the agent's capabilities at runtime

func (*EnhancedAgent) WithdrawPublic added in v0.6.5 

func (a *EnhancedAgent) WithdrawPublic() error

WithdrawPublic withdraws a public agent back to private visibility. Only works on agents that are currently public.

type EnhancedAgentConfig 

type EnhancedAgentConfig struct {
	Config       *Config
	AgentHandler types.AgentHandler

	// NFT Minting Options (choose one: Deploy, Mint, or provide TokenID)
	Deploy  bool   // If true, use new secure deploy flow with database persistence
	Mint    bool   // If true, use legacy mint flow (no database persistence)
	TokenID uint64 // Required if Deploy and Mint are both false

	// Deploy-specific options
	AgentID       string // Required for Deploy, auto-generated from name if empty
	AgentType     string // Agent type: "command", "nlp", "mcp", "commandless" (default: "command")
	StateFilePath string // Path to state file for Deploy (default: .teneo-deploy-state.json)

	// Backend Configuration
	BackendURL  string // Default from env or "http://localhost:8080"
	RPCEndpoint string // Ethereum RPC endpoint

	// Optional: Submit agent for public visibility review after startup (defaults to false).
	// The agent goes through a review process (up to 72 hours) before becoming publicly visible.
	SubmitForReview bool
}

EnhancedAgentConfig represents configuration for the enhanced agent

type OpenAIAgent 

type OpenAIAgent struct {
	// contains filtered or unexported fields
}

OpenAIAgent wraps the OpenAI client to implement the AgentHandler interface

func NewOpenAIAgent 

func NewOpenAIAgent(config *OpenAIConfig) *OpenAIAgent

NewOpenAIAgent creates a new OpenAI-powered agent handler

func (*OpenAIAgent) IsStreaming 

func (a *OpenAIAgent) IsStreaming() bool

IsStreaming returns whether streaming is enabled

func (*OpenAIAgent) ProcessTask 

func (a *OpenAIAgent) ProcessTask(ctx context.Context, task string) (string, error)

ProcessTask implements the AgentHandler interface

func (*OpenAIAgent) ProcessTaskWithStreaming 

func (a *OpenAIAgent) ProcessTaskWithStreaming(ctx context.Context, task string, sender types.MessageSender) error

ProcessTaskWithStreaming implements the StreamingTaskHandler interface This method is called by the SDK if the agent implements StreamingTaskHandler. If streaming is disabled, it falls back to ProcessTask and sends a single message.

func (*OpenAIAgent) SetMaxTokens 

func (a *OpenAIAgent) SetMaxTokens(tokens int)

SetMaxTokens updates the max tokens

func (*OpenAIAgent) SetStreaming 

func (a *OpenAIAgent) SetStreaming(enabled bool)

SetStreaming enables or disables streaming responses

func (*OpenAIAgent) SetSystemPrompt 

func (a *OpenAIAgent) SetSystemPrompt(prompt string)

SetSystemPrompt updates the system prompt

func (*OpenAIAgent) SetTemperature 

func (a *OpenAIAgent) SetTemperature(temp float32)

SetTemperature updates the temperature

type OpenAIConfig 

type OpenAIConfig struct {
	APIKey       string  // OpenAI API key
	Model        string  // Model to use (e.g., "gpt-5", "gpt-4", "gpt-3.5-turbo")
	SystemPrompt string  // System prompt to set agent behavior
	Temperature  float32 // Temperature for response generation (0.0 - 2.0). Note: Beta models (GPT-5, O1, O3) have fixed temperature=1
	MaxTokens    int     // Maximum tokens in response
	Streaming    bool    // Enable streaming responses (default: false)
}

OpenAIConfig holds configuration for the OpenAI agent

type SimpleOpenAIAgentConfig 

type SimpleOpenAIAgentConfig struct {
	// Required: Your Ethereum private key for Teneo network authentication
	PrivateKey string

	// Required: Your OpenAI API key
	OpenAIKey string

	// Optional: Agent name (defaults to "OpenAI Agent")
	Name string

	// Optional: Agent description
	Description string

	// Optional: OpenAI model (defaults to "gpt-5")
	Model string

	// Optional: System prompt for the AI (defaults to helpful assistant)
	SystemPrompt string

	// Optional: Temperature 0.0-2.0 (defaults to 0.7)
	// Note: Beta models (GPT-5, O1, O3) have fixed temperature=1 and will ignore this setting
	Temperature float32

	// Optional: Max tokens per response (defaults to 1000)
	MaxTokens int

	// Optional: Enable streaming responses (defaults to false - single message)
	Streaming bool

	// Optional: Agent capabilities (defaults to ["chat", "text_generation"])
	Capabilities []string

	// Optional: NFT Token ID (if you already have one, otherwise set Mint to true)
	TokenID uint64

	// Optional: Mint new NFT (defaults to false)
	Mint bool

	// Optional: Submit agent for public visibility review after startup (defaults to false).
	// The agent goes through a review process (up to 72 hours) before becoming publicly visible.
	// The agent must stay online during review.
	SubmitForReview bool

	// Optional: WebSocket URL (defaults to env WEBSOCKET_URL or standard endpoint)
	WebSocketURL string

	// Optional: Room to join (defaults to empty string)
	Room string

	// Optional: Rate limit per minute (defaults to 0 = unlimited)
	RateLimitPerMinute int

	// Optional: Task timeout in seconds (defaults to 120s for beta models like GPT-5/O1/O3, 30s for others)
	TaskTimeout int
}

SimpleOpenAIAgentConfig provides a minimal configuration for quick OpenAI agent setup

type SubmitForReviewResult added in v0.8.2 

type SubmitForReviewResult struct {
	Success           bool   `json:"success"`
	Status            string `json:"status,omitempty"`
	Message           string `json:"message,omitempty"`
	ReviewStatus      string `json:"review_status,omitempty"`
	ReviewStateBefore string `json:"review_state_before,omitempty"`
	ReviewStateAfter  string `json:"review_state_after,omitempty"`
	HashBefore        string `json:"hash_before,omitempty"`
	HashAfter         string `json:"hash_after,omitempty"`
	SubmittedAt       string `json:"submitted_at,omitempty"`
	Error             string `json:"error,omitempty"`
}

func SubmitForReviewDetailed added in v0.8.2 

func SubmitForReviewDetailed(backendURL, agentID, creatorWallet string, tokenID uint64, headers ...map[string]string) (*SubmitForReviewResult, error)

SubmitForReview submits an agent for public visibility review on the Teneo network.

Agents go through a review process before becoming publicly visible: 

private → in_review → public (approved) or declined

This is a standalone utility that can be called from any context — no running agent required. The agent must have been deployed, connected at least once, and be currently online.

Parameters:

  • backendURL: The Teneo backend URL (e.g. "https://backend.developer.chatroom.teneo-protocol.ai")
  • agentID: The agent's ID from your metadata JSON (agentId field)
  • creatorWallet: The Ethereum wallet address that owns the agent's NFT
  • tokenID: The NFT token ID for on-chain ownership verification

Example: 

result, err := agent.SubmitForReview(
    "https://backend.developer.chatroom.teneo-protocol.ai",
    "my-agent-id",
    "0xYourWalletAddress",
    42,
)
fmt.Println(result.Status)

HTTP API equivalent (for non-Go clients): 

POST {backendURL}/api/agents/{agent-id}/submit-for-review
Content-Type: application/json

{
    "creator_wallet": "0xYourWalletAddress",
    "token_id": 42
}

Source Files

View all Source files

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.