cli

package
v0.7.2 Latest Latest
Warning

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

 Go to latest Published: Jun 17, 2026 License: Apache-2.0 Imports: 33 Imported by: 0

Documentation

Overview

Package cli provides sophisticated command-line interface utilities for the muster application.

This package offers a comprehensive CLI experience with intelligent data formatting, robust server connectivity, and professional output presentation. It serves as the primary interface layer between user commands and the muster aggregator system.

Core Components

ToolExecutor provides high-level tool execution with multiple output formats:

  • MCP (Message Control Protocol) client integration with the aggregator
  • Multiple output formats: professional tables, JSON, and YAML
  • Progress indicators with spinners for long-running operations
  • Server connectivity validation and health checking
  • Both formatted and programmatic execution modes

TableFormatter offers intelligent table creation and optimization:

  • Auto-detection of muster resource types (services, workflows, capabilities, etc.)
  • Smart column selection and optimization based on data structure
  • Specialized formatting rules for different resource types
  • Professional styling with consistent iconography and color coding
  • Handles both simple arrays and complex nested object structures

TableBuilder provides cell-level formatting and styling utilities:

  • Status indicators with color-coded icons (✅ healthy, ❌ unhealthy, etc.)
  • Resource-specific formatting (tool names, metadata, descriptions)
  • Consistent styling for names, IDs, types, and other common fields
  • Smart truncation and summarization of complex data
  • Icon assignment based on resource types

Common utilities handle server connectivity and message formatting:

  • Auto-detection of aggregator endpoints from configuration
  • Server health checks with appropriate error messages
  • Consistent formatting for success (✓), error, and warning (⚠) messages
  • Configuration-aware endpoint resolution

Output Formats

The package supports three output formats to accommodate different use cases:

  • Table: Professional tables with optimized columns, styling, and summaries
  • JSON: Raw JSON output for programmatic consumption
  • YAML: Human-readable YAML format converted from JSON responses

Resource Type Intelligence

The formatter automatically detects and optimizes display for muster resource types:

  • Services: Health status, state, service type, and metadata
  • MCP Servers: Server state, type, and configuration details
  • Workflows: Steps, input args, and execution details

Usage Patterns

The package is designed for use by muster CLI commands that need to:

  • Execute tools through the aggregator with professional output
  • Display complex data structures in user-friendly formats
  • Provide consistent styling and iconography across all commands
  • Validate server connectivity before attempting operations
  • Support both interactive and programmatic usage scenarios

Integration

All CLI commands should use this package for consistent user experience. The package handles the complexity of MCP communication, data formatting, and output presentation, allowing commands to focus on their core logic.

Index

Constants

View Source 
const (
	// ServerTypeStdio represents a local stdio-based MCP server.
	ServerTypeStdio = "stdio"
	// ServerTypeStreamableHTTP represents a remote streamable HTTP MCP server.
	ServerTypeStreamableHTTP = "streamable-http"
	// ServerTypeSSE represents a remote Server-Sent Events MCP server.
	ServerTypeSSE = "sse"
)

Server type constants for MCP server classification.

View Source 
const AuthModeEnvVar = "MUSTER_AUTH_MODE"

AuthModeEnvVar is the environment variable name for setting the default auth mode.

View Source 
const CallbackPortEnvVar = "MUSTER_OAUTH_CALLBACK_PORT"

CallbackPortEnvVar is the environment variable for configuring the OAuth callback port.

View Source 
const ContextEnvVar = musterctx.ContextEnvVar

ContextEnvVar is the environment variable name for overriding the current context.

View Source 
const DefaultCallbackPort = 3000

DefaultCallbackPort is the default port for OAuth callbacks if not configured.

View Source 
const DefaultConnectionCheckTimeout = 5 * time.Second

DefaultConnectionCheckTimeout is the timeout for checking connection state and tokens.

View Source 
const DefaultHTTPClientTimeout = 5 * time.Second

DefaultHTTPClientTimeout is the timeout for HTTP client operations.

View Source 
const EndpointEnvVar = "MUSTER_ENDPOINT"

EndpointEnvVar is the environment variable name for setting the default endpoint.

Variables

View Source 
var ValidOutputFormats = []OutputFormat{
	OutputFormatTable,
	OutputFormatWide,
	OutputFormatJSON,
	OutputFormatYAML,
}

ValidOutputFormats contains all valid output format values.

Functions

func CheckServerRunning 

func CheckServerRunning(endpoint string) error

CheckServerRunning verifies that the muster aggregator server is running and responsive. It performs a health check by sending a GET request to the MCP endpoint and validates the response status code. This is typically used before attempting to execute commands that require server connectivity.

Returns:

  • error: nil if server is running and responsive, otherwise an error with guidance

func ExtractServerType 

func ExtractServerType(data map[string]interface{}) string

ExtractServerType extracts the server type from a data map. It checks common field names that indicate the server type in MCP server data.

The function checks the following fields in order:

  • "type": Direct server type field
  • "metadata": May contain server type as a string

Args:

  • data: Map containing server data

Returns:

  • string: The server type (stdio, streamable-http, sse) or empty string if not found

func FormatMCPPromptDetail 

func FormatMCPPromptDetail(prompt MCPPrompt, format OutputFormat) error

FormatMCPPromptDetail formats and displays detailed MCP prompt info. For table format, it uses a kubectl-describe-like plain text output.

func FormatMCPPromptsWithOptions 

func FormatMCPPromptsWithOptions(prompts []MCPPrompt, format OutputFormat, noHeaders bool) error

FormatMCPPromptsWithOptions formats and displays MCP prompts with additional options.

func FormatMCPResourceDetail 

func FormatMCPResourceDetail(resource MCPResource, format OutputFormat) error

FormatMCPResourceDetail formats and displays detailed MCP resource info. For table format, it uses a kubectl-describe-like plain text output.

func FormatMCPResourcesWithOptions 

func FormatMCPResourcesWithOptions(resources []MCPResource, format OutputFormat, noHeaders bool) error

FormatMCPResourcesWithOptions formats and displays MCP resources with additional options.

func FormatMCPToolDetail 

func FormatMCPToolDetail(tool MCPTool, format OutputFormat) error

FormatMCPToolDetail formats and displays detailed MCP tool info. For table format, it uses a kubectl-describe-like plain text output.

func FormatMCPToolsWithOptions 

func FormatMCPToolsWithOptions(tools []MCPTool, format OutputFormat, noHeaders bool) error

FormatMCPToolsWithOptions formats and displays MCP tools with additional options.

func GetAggregatorEndpoint 

func GetAggregatorEndpoint(cfg *config.MusterConfig) string

GetAggregatorEndpoint detects and returns the aggregator endpoint URL from a specific configuration.

Args:

  • cfg: Pointer to MusterConfig to use for detecting the endpoint.

Returns:

  • string: The complete HTTP endpoint URL (e.g., "http://localhost:8090/mcp")
  • error: Always nil (kept for future compatibility)

func GetDefaultEndpoint 

func GetDefaultEndpoint() string

GetDefaultEndpoint returns the endpoint from environment variable if set.

func IsMCPServerData 

func IsMCPServerData(data map[string]interface{}) bool

IsMCPServerData checks if the data represents an MCP server. It analyzes the data structure to identify MCP server-specific patterns.

Args:

  • data: Map containing potential MCP server data

Returns:

  • bool: true if the data appears to be MCP server related

func IsRemoteEndpoint 

func IsRemoteEndpoint(endpoint string) bool

IsRemoteEndpoint checks if an endpoint URL points to a remote server. It properly parses the URL and checks only the hostname, avoiding false positives when "localhost" appears in the path or query string.

Args:

  • endpoint: The endpoint URL to check

Returns:

  • bool: true if the endpoint is remote (not localhost), false otherwise

func IsRemoteServerType 

func IsRemoteServerType(serverType string) bool

IsRemoteServerType returns true if the server type represents a remote connection. Remote servers use network connections (streamable-http or sse) rather than local process communication (stdio).

Args:

  • serverType: The server type string to check

Returns:

  • bool: true if the server type is remote (streamable-http or sse)

func ParseTimeFilter 

func ParseTimeFilter(timeStr string) (time.Time, error)

ParseTimeFilter parses time strings in various formats for event filtering and other time-based queries. It supports duration format (e.g., "1h", "30m"), RFC3339 format (e.g., "2024-01-15T10:00:00Z"), date-only format (e.g., "2024-01-15"), and datetime format (e.g., "2024-01-15 10:00:00").

For duration format, the returned time is calculated as now minus the duration.

Args:

  • timeStr: The time string to parse

Returns:

  • time.Time: The parsed time
  • error: An error if the format is not recognized

func RegisterCommonFlags 

func RegisterCommonFlags(cmd *cobra.Command, flags *CommandFlags)

RegisterCommonFlags registers the common flags used by most CLI commands that connect to a muster aggregator. This reduces duplication across command files and ensures consistent flag naming and descriptions.

The registered flags are:

  • --output/-o: Output format (table, wide, json, yaml), default: "table"
  • --no-headers: Suppress header row in table output
  • --quiet/-q: Suppress non-essential output
  • --debug: Enable debug logging (show MCP protocol messages)
  • --config-path: Configuration directory
  • --endpoint: Remote muster aggregator endpoint URL (env: MUSTER_ENDPOINT)
  • --context: Use a specific context (env: MUSTER_CONTEXT)
  • --auth: Authentication mode (env: MUSTER_AUTH_MODE)

func ResolveEndpoint 

func ResolveEndpoint(explicitEndpoint, contextName string) (string, error)

ResolveEndpoint resolves the endpoint URL using the precedence order: 1. Explicit endpoint (from --endpoint flag) 2. Context name (from --context flag) 3. MUSTER_CONTEXT environment variable 4. current-context from contexts.yaml 5. Empty string (caller should fall back to config-based endpoint)

Returns the resolved endpoint URL and any error encountered. If no context is configured, returns an empty string (not an error).

func ValidateOutputFormat 

func ValidateOutputFormat(format string) error

ValidateOutputFormat validates that the given format string is a supported output format. Returns nil if valid, or an error with a helpful message listing valid formats.

Types

type AuthAdapter 

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

AuthAdapter implements api.AuthHandler using internal/agent/oauth. It wraps the AuthManager and TokenStore to provide OAuth authentication for CLI commands following the project's service locator pattern.

Thread-safe: All public methods are safe for concurrent use.

func NewAuthAdapter 

func NewAuthAdapter() (*AuthAdapter, error)

NewAuthAdapter creates a new auth adapter with default configuration.

func NewAuthAdapterWithConfig 

func NewAuthAdapterWithConfig(cfg AuthAdapterConfig) (*AuthAdapter, error)

NewAuthAdapterWithConfig creates a new auth adapter with the specified configuration. This is useful for testing or advanced use cases where custom token storage is needed.

func (*AuthAdapter) CheckAuthRequired 

func (a *AuthAdapter) CheckAuthRequired(ctx context.Context, endpoint string) (bool, error)

CheckAuthRequired probes the endpoint to determine if OAuth is required.

func (*AuthAdapter) Close 

func (a *AuthAdapter) Close() error

Close cleans up any resources held by the auth adapter.

func (*AuthAdapter) GetBearerToken 

func (a *AuthAdapter) GetBearerToken(endpoint string) (string, error)

GetBearerToken returns a valid Bearer token for the endpoint. Token refresh is handled by mcp-go's transport layer via AgentTokenStore.

func (*AuthAdapter) GetStatus 

func (a *AuthAdapter) GetStatus() []api.AuthStatus

GetStatus returns authentication status for all known endpoints.

func (*AuthAdapter) GetStatusForEndpoint 

func (a *AuthAdapter) GetStatusForEndpoint(endpoint string) *api.AuthStatus

GetStatusForEndpoint returns authentication status for a specific endpoint.

func (*AuthAdapter) HasCredentials added in v0.1.1 

func (a *AuthAdapter) HasCredentials(endpoint string) bool

HasCredentials reports whether usable credentials exist for the endpoint: a valid access token or an expired token with a refresh token.

func (*AuthAdapter) InvalidateCache added in v0.1.1 

func (a *AuthAdapter) InvalidateCache(endpoint string)

InvalidateCache removes the cached auth manager for an endpoint. This forces the next GetStatusForEndpoint call to create a fresh manager that reads the latest token from the file store. This is needed after mcp-go's transport refreshes a token, since the refreshed token is persisted to file by AgentTokenStore but the AuthAdapter's in-memory TokenStore cache is stale.

func (*AuthAdapter) Login 

func (a *AuthAdapter) Login(ctx context.Context, endpoint string) error

Login initiates the OAuth flow for the given endpoint. If a previous session exists, it first attempts silent re-authentication using prompt=none. If silent auth fails, it falls back to interactive login.

func (*AuthAdapter) LoginWithIssuer 

func (a *AuthAdapter) LoginWithIssuer(ctx context.Context, endpoint, issuerURL string) error

LoginWithIssuer initiates the OAuth flow with a known issuer.

func (*AuthAdapter) Logout 

func (a *AuthAdapter) Logout(endpoint string) error

Logout clears stored tokens for the endpoint. It first revokes the refresh token via RFC 7009 (best-effort), then performs local cleanup.

func (*AuthAdapter) LogoutAll 

func (a *AuthAdapter) LogoutAll() error

LogoutAll clears all stored tokens. It revokes each endpoint's refresh token via RFC 7009, then calls DELETE /user-tokens to clear server-side downstream state.

func (*AuthAdapter) Register 

func (a *AuthAdapter) Register()

Register registers the adapter with the API layer.

func (*AuthAdapter) SetNoSilentRefresh 

func (a *AuthAdapter) SetNoSilentRefresh(noSilent bool)

SetNoSilentRefresh enables or disables silent re-authentication. When disabled (the default), Login() attempts silent re-auth before interactive login. When enabled, Login() always uses interactive authentication.

type AuthAdapterConfig 

type AuthAdapterConfig struct {
	// TokenStorageDir is the directory for storing OAuth tokens.
	// If empty, defaults to ~/.config/muster/tokens
	TokenStorageDir string

	// NoSilentRefresh disables silent re-authentication attempts.
	// When true, Login() always uses interactive authentication.
	NoSilentRefresh bool
}

AuthAdapterConfig provides configuration options for the AuthAdapter.

type AuthExpiredError 

type AuthExpiredError struct {
	// Endpoint is the URL whose token has expired.
	Endpoint string
}

AuthExpiredError indicates the token has expired. Implements error with actionable guidance.

func (*AuthExpiredError) Error 

func (e *AuthExpiredError) Error() string

Error returns a user-friendly error message with actionable guidance.

func (*AuthExpiredError) Is 

func (e *AuthExpiredError) Is(target error) bool

Is allows errors.Is() to work with wrapped errors.

type AuthFailedError 

type AuthFailedError struct {
	// Endpoint is the URL where authentication failed.
	Endpoint string
	// Reason is the underlying error.
	Reason error
}

AuthFailedError indicates authentication failed.

func (*AuthFailedError) Error 

func (e *AuthFailedError) Error() string

Error returns a user-friendly error message with actionable guidance.

func (*AuthFailedError) Is 

func (e *AuthFailedError) Is(target error) bool

Is allows errors.Is() to work with wrapped errors.

func (*AuthFailedError) Unwrap 

func (e *AuthFailedError) Unwrap() error

Unwrap returns the underlying error.

type AuthMode 

type AuthMode string

AuthMode represents authentication behavior for CLI commands. 

const (
	// AuthModeAuto automatically triggers OAuth browser login when authentication is required.
	// This is the default behavior.
	AuthModeAuto AuthMode = "auto"
	// AuthModePrompt prompts the user before triggering authentication.
	AuthModePrompt AuthMode = "prompt"
	// AuthModeNone fails immediately on 401 without attempting authentication.
	AuthModeNone AuthMode = "none"
)

func GetAuthModeWithOverride 

func GetAuthModeWithOverride(override string) (AuthMode, error)

GetAuthModeWithOverride returns the auth mode from the provided override string, falling back to the environment variable default if the override is empty. This consolidates the common pattern used across CLI commands.

Note: ParseAuthMode already handles empty string as "auto", so this function adds environment variable lookup as an intermediate step.

func GetDefaultAuthMode 

func GetDefaultAuthMode() AuthMode

GetDefaultAuthMode returns the default auth mode from environment or "auto".

func ParseAuthMode 

func ParseAuthMode(s string) (AuthMode, error)

ParseAuthMode parses a string into an AuthMode, with validation.

type AuthRequiredError 

type AuthRequiredError struct {
	// Endpoint is the URL that requires authentication.
	Endpoint string
}

AuthRequiredError indicates authentication is needed. Implements error with actionable guidance.

func (*AuthRequiredError) Error 

func (e *AuthRequiredError) Error() string

Error returns a user-friendly error message with actionable guidance.

func (*AuthRequiredError) Is 

func (e *AuthRequiredError) Is(target error) bool

Is allows errors.Is() to work with wrapped errors.

type CommandFlags 

type CommandFlags struct {
	// OutputFormat specifies the desired output format (table, json, yaml)
	OutputFormat string
	// NoHeaders suppresses the header row in table output
	NoHeaders bool
	// Quiet suppresses progress indicators and non-essential output
	Quiet bool
	// Debug enables verbose logging of MCP protocol messages
	Debug bool
	// ConfigPath specifies a custom configuration directory path
	ConfigPath string
	// Endpoint overrides the aggregator endpoint URL for remote connections
	Endpoint string
	// Context specifies a named context to use for endpoint resolution
	Context string
	// AuthMode controls authentication behavior (auto, prompt, none)
	AuthMode string
}

CommandFlags holds the common flag values used across CLI commands that connect to a muster aggregator. This struct consolidates the repetitive flag pattern used by commands like get, list, check, start, stop, create, and events.

func (*CommandFlags) ToExecutorOptions 

func (f *CommandFlags) ToExecutorOptions() (ExecutorOptions, error)

ToExecutorOptions converts CommandFlags to ExecutorOptions for use with NewToolExecutor. This provides a convenient bridge between the flag registration and executor creation. It validates the output format and returns an error for unsupported formats.

type ConnectionError 

type ConnectionError struct {
	// Endpoint is the URL that could not be reached.
	Endpoint string
	// Type categorizes the connection error.
	Type ConnectionErrorType
	// Reason is the underlying error.
	Reason error
}

ConnectionError indicates a connection failure to an endpoint. It wraps the underlying error and provides categorization for better user feedback.

func ClassifyConnectionError 

func ClassifyConnectionError(err error, endpoint string) *ConnectionError

ClassifyConnectionError analyzes an error and returns a ConnectionError with the appropriate type. If the error is nil, returns nil.

type ConnectionErrorType 

type ConnectionErrorType int

ConnectionErrorType categorizes the type of connection error. 

const (
	// ConnectionErrorUnknown indicates an unclassified connection error.
	ConnectionErrorUnknown ConnectionErrorType = iota
	// ConnectionErrorTLS indicates a TLS/certificate verification error.
	ConnectionErrorTLS
	// ConnectionErrorNetwork indicates a network connectivity error (e.g., refused, unreachable).
	ConnectionErrorNetwork
	// ConnectionErrorTimeout indicates a connection timeout.
	ConnectionErrorTimeout
	// ConnectionErrorDNS indicates a DNS resolution failure.
	ConnectionErrorDNS
)

func (ConnectionErrorType) String 

func (t ConnectionErrorType) String() string

String returns a human-readable name for the connection error type.

type ExecutorOptions 

type ExecutorOptions struct {
	// Format specifies the desired output format (table, json, yaml)
	Format OutputFormat
	// NoHeaders suppresses the header row in table output
	NoHeaders bool
	// Quiet suppresses progress indicators and non-essential output
	Quiet bool
	// Debug enables verbose logging of MCP protocol messages and initialization
	Debug bool
	// ConfigPath specifies a custom configuration directory path
	ConfigPath string
	// Endpoint overrides the aggregator endpoint URL for remote connections
	Endpoint string
	// Context specifies a named context to use for endpoint resolution
	Context string
	// AuthMode controls authentication behavior (auto, prompt, none)
	AuthMode AuthMode
}

ExecutorOptions contains configuration options for tool execution. These options control how commands are executed and how output is formatted.

type MCPPrompt 

type MCPPrompt = mcp.Prompt

MCPPrompt is an alias for mcp.Prompt for use in cmd package

type MCPResource 

type MCPResource = mcp.Resource

MCPResource is an alias for mcp.Resource for use in cmd package

type MCPTool 

type MCPTool = mcp.Tool

MCPTool is an alias for mcp.Tool for use in cmd package

type OutputFormat 

type OutputFormat string

OutputFormat represents the supported output formats for CLI commands. This allows users to choose how they want to receive command results. 

const (
	// OutputFormatTable formats output as a kubectl-style plain table
	OutputFormatTable OutputFormat = "table"
	// OutputFormatWide formats output as a table with additional columns
	OutputFormatWide OutputFormat = "wide"
	// OutputFormatJSON formats output as raw JSON data
	OutputFormatJSON OutputFormat = "json"
	// OutputFormatYAML formats output as YAML data converted from JSON
	OutputFormatYAML OutputFormat = "yaml"
)

type PlainTableWriter 

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

PlainTableWriter provides kubectl-style plain table output without box-drawing characters. This format is optimized for:

  • Easy copy/paste operations
  • Piping to grep, awk, cut and other command-line tools
  • Terminal-agnostic rendering (no Unicode issues)
  • Familiar kubectl-like appearance

func NewPlainTableWriter 

func NewPlainTableWriter(output io.Writer) *PlainTableWriter

NewPlainTableWriter creates a new plain table writer with kubectl-style formatting. By default, headers are shown. Use SetNoHeaders(true) to suppress them.

func (*PlainTableWriter) AppendRow 

func (w *PlainTableWriter) AppendRow(row []string)

AppendRow adds a row to the table.

func (*PlainTableWriter) Render 

func (w *PlainTableWriter) Render()

Render outputs the table in kubectl-style format.

func (*PlainTableWriter) SetHeaders 

func (w *PlainTableWriter) SetHeaders(headers []string)

SetHeaders sets the column headers for the table. Headers are displayed in uppercase by default.

func (*PlainTableWriter) SetNoHeaders 

func (w *PlainTableWriter) SetNoHeaders(noHeaders bool)

SetNoHeaders controls whether to suppress the header row.

type ServerStatus 

type ServerStatus struct {
	// Endpoint is the URL of the server.
	Endpoint string

	// Reachable indicates whether the server responded to probes.
	Reachable bool

	// AuthRequired indicates whether the server requires authentication.
	AuthRequired bool

	// Authenticated indicates whether we have a valid token for this server.
	Authenticated bool

	// Error holds any error that occurred during the check.
	Error error
}

ServerStatus represents the connectivity and auth status of a server.

type TableBuilder 

type TableBuilder struct{}

TableBuilder handles cell formatting and styling for table display. It provides specialized formatting for different types of data commonly encountered in muster operations, including status indicators, metadata, and resource-specific information. The builder applies consistent styling and iconography to create professional-looking table output.

func NewTableBuilder 

func NewTableBuilder() *TableBuilder

NewTableBuilder creates a new table builder instance. The builder is stateless and can be reused for multiple formatting operations.

Returns:

  • *TableBuilder: New table builder ready for use

func (*TableBuilder) FormatCellValuePlain 

func (b *TableBuilder) FormatCellValuePlain(column string, value interface{}, rowContext map[string]interface{}) string

FormatCellValuePlain formats individual cell values as plain text without ANSI styling. This is used for kubectl-style table output where we don't want color codes.

Args:

  • column: The column name/type to determine formatting rules
  • value: The raw value to format
  • rowContext: The full row data for context-aware formatting (may be nil)

Returns:

  • string: Formatted value as plain text

func (*TableBuilder) SimplifyToolName 

func (b *TableBuilder) SimplifyToolName(toolName string) string

SimplifyToolName removes common prefixes to make tool names more readable. Many muster tools have predictable prefixes that add noise to table display. This function strips common prefixes to improve readability.

Args:

  • toolName: The full tool name to simplify

Returns:

  • string: Simplified tool name with common prefixes removed

func (*TableBuilder) SortDataByName 

func (b *TableBuilder) SortDataByName(data []interface{}, columns []string) []interface{}

SortDataByName sorts data by the first column (usually name/id). This provides consistent ordering in tables, making it easier for users to find specific resources.

Args:

  • data: Array of data objects to sort
  • columns: Column names, with the first used for sorting

Returns:

  • []interface{}: Sorted data array

type TableFormatter 

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

TableFormatter handles table creation and optimization for muster CLI output. It provides intelligent formatting for different types of data structures, automatically optimizing column layouts and applying consistent styling. The formatter can handle both simple arrays and complex nested objects, adapting the display format to provide the best user experience.

func NewTableFormatter 

func NewTableFormatter(options ExecutorOptions) *TableFormatter

NewTableFormatter creates a new table formatter with the specified options. The formatter uses the provided options to determine output behavior, including format selection and verbosity settings.

Args:

  • options: Configuration options for formatting behavior

Returns:

  • *TableFormatter: Configured table formatter ready for use

func (*TableFormatter) FormatData 

func (f *TableFormatter) FormatData(data interface{}) error

FormatData formats data according to its type and structure. It intelligently handles different data types including objects, arrays, and simple values, applying the most appropriate formatting strategy for optimal readability and information density.

Args:

  • data: The data to format (can be object, array, or simple value)

Returns:

  • error: Formatting error, if any

type ToolExecutor 

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

ToolExecutor provides high-level tool execution functionality with formatted output. It handles the connection to the muster aggregator, executes tools, and formats the results according to the specified output format. This is the main interface for CLI commands that need to interact with muster services.

func NewToolExecutor 

func NewToolExecutor(options ExecutorOptions) (*ToolExecutor, error)

NewToolExecutor creates a new tool executor with the specified options. It establishes the connection configuration and initializes the MCP client for communication with the muster aggregator server.

Args:

  • options: Configuration options for execution and output formatting

Returns:

  • *ToolExecutor: Configured tool executor ready for use
  • error: Configuration or connection setup error

func (*ToolExecutor) Close 

func (e *ToolExecutor) Close() error

Close gracefully closes the connection to the aggregator server. This should be called when the executor is no longer needed to free resources.

Returns:

  • error: Error during connection cleanup, if any

func (*ToolExecutor) Connect 

func (e *ToolExecutor) Connect(ctx context.Context) error

Connect establishes a connection to the muster aggregator server. It shows a progress spinner unless quiet mode is enabled, and handles connection errors with appropriate user feedback. For remote servers, it uses mcp-go's OAuth transport for automatic token injection and typed 401 error handling.

Args:

  • ctx: Context for connection timeout and cancellation

Returns:

  • error: Connection error, if any

func (*ToolExecutor) Execute 

func (e *ToolExecutor) Execute(ctx context.Context, toolName string, args map[string]interface{}) error

Execute executes a tool with the given args and formats the output. This is the main method for running muster tools through the CLI interface. It handles progress indication, error formatting, and output formatting according to the configured options.

Args:

  • ctx: Context for execution timeout and cancellation
  • toolName: Name of the tool to execute
  • args: Tool args as key-value pairs

Returns:

  • error: Execution or formatting error, if any

func (*ToolExecutor) ExecuteJSON 

func (e *ToolExecutor) ExecuteJSON(ctx context.Context, toolName string, args map[string]interface{}) (interface{}, error)

ExecuteJSON executes a tool and returns the result as parsed JSON. This method is useful when you need to work with structured data programmatically rather than displaying it to users.

Args:

  • ctx: Context for execution timeout and cancellation
  • toolName: Name of the tool to execute
  • args: Tool args as key-value pairs

Returns:

  • interface{}: Parsed JSON result as Go data structures
  • error: Execution or JSON parsing error, if any

func (*ToolExecutor) GetClient 

func (e *ToolExecutor) GetClient() *agent.Client

GetClient returns the underlying agent client for advanced use cases like streaming.

func (*ToolExecutor) GetMCPPrompt 

func (e *ToolExecutor) GetMCPPrompt(ctx context.Context, name string) (*mcp.Prompt, error)

GetMCPPrompt returns detailed info for a specific prompt. This method retrieves the prompt list and finds the specified prompt by name.

Args:

  • ctx: Context for execution timeout and cancellation
  • name: The exact name of the prompt to find

Returns:

  • *mcp.Prompt: Pointer to the found prompt, or nil if not found
  • error: Connection or retrieval error, if any

func (*ToolExecutor) GetMCPResource 

func (e *ToolExecutor) GetMCPResource(ctx context.Context, uri string) (*mcp.Resource, error)

GetMCPResource returns detailed info for a specific resource. This method retrieves the resource list and finds the specified resource by URI.

Args:

  • ctx: Context for execution timeout and cancellation
  • uri: The exact URI of the resource to find

Returns:

  • *mcp.Resource: Pointer to the found resource, or nil if not found
  • error: Connection or retrieval error, if any

func (*ToolExecutor) GetMCPTool 

func (e *ToolExecutor) GetMCPTool(ctx context.Context, name string) (*mcp.Tool, error)

GetMCPTool returns detailed info for a specific tool by calling describe_tool meta-tool. This method retrieves the tool information from the actual tools (core_*, x_*, workflow_*) rather than the meta-tools exposed by the MCP native protocol.

Args:

  • ctx: Context for execution timeout and cancellation
  • name: The exact name of the tool to find

Returns:

  • *mcp.Tool: Pointer to the found tool, or nil if not found
  • error: Connection or retrieval error, if any

func (*ToolExecutor) GetOptions 

func (e *ToolExecutor) GetOptions() ExecutorOptions

GetOptions returns the executor options. This allows callers to check the configured output format and other settings.

func (*ToolExecutor) ListMCPPrompts 

func (e *ToolExecutor) ListMCPPrompts(ctx context.Context) ([]mcp.Prompt, error)

ListMCPPrompts returns all MCP prompts using native protocol. This method retrieves prompts directly from the MCP server without going through the tool execution interface.

Args:

  • ctx: Context for execution timeout and cancellation

Returns:

  • []mcp.Prompt: Slice of all available prompts from the server
  • error: Connection or retrieval error, if any

func (*ToolExecutor) ListMCPResources 

func (e *ToolExecutor) ListMCPResources(ctx context.Context) ([]mcp.Resource, error)

ListMCPResources returns all MCP resources using native protocol. This method retrieves resources directly from the MCP server without going through the tool execution interface.

Args:

  • ctx: Context for execution timeout and cancellation

Returns:

  • []mcp.Resource: Slice of all available resources from the server
  • error: Connection or retrieval error, if any

func (*ToolExecutor) ListMCPTools 

func (e *ToolExecutor) ListMCPTools(ctx context.Context) ([]mcp.Tool, error)

ListMCPTools returns all MCP tools by calling the list_tools meta-tool. This method retrieves the actual tools (core_*, x_*, workflow_*) rather than the meta-tools exposed by the MCP native tools/list protocol.

Args:

  • ctx: Context for execution timeout and cancellation

Returns:

  • []mcp.Tool: Slice of all available tools from the server
  • error: Connection or retrieval error, if any

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.