Documentation
¶
Overview ¶
Package schema provides access to Lacquer workflow schema definitions and metadata. This package enables third-party applications to introspect Lacquer's capabilities, including the workflow schema structure, available expression syntax, built-in functions, and supported AI model providers.
The schema information is essential for:
- Building workflow editors and IDEs with syntax validation
- Creating workflow validation tools
- Generating documentation for workflow syntax
- Discovering available AI models and providers
- Understanding supported expression and function syntax
Example usage:
// Get the complete schema information
schema, err := GetSchema()
if err != nil {
log.Fatal(err)
}
// Access the JSON schema for workflow validation
var workflowSchema map[string]interface{}
json.Unmarshal(schema.Schema, &workflowSchema)
// List available AI models
for _, provider := range schema.ModelProviders {
fmt.Printf("Provider: %s\n", provider.Provider)
for _, model := range provider.Models {
fmt.Printf(" - %s\n", model)
}
}
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type Model ¶
type Model struct {
// Provider is the name/identifier of the AI model provider.
// Common providers include "openai", "anthropic", and "local".
Provider string `json:"provider"`
// Models is a list of model identifiers available from this provider.
// These IDs can be used directly in workflow agent configurations
// to specify which model should handle the agent's tasks.
//
// Examples: ["gpt-4", "gpt-3.5-turbo"] for OpenAI,
// ["claude-3-opus", "claude-3-sonnet"] for Anthropic.
Models []string `json:"models"`
}
Model represents an AI model provider and its available models. This structure provides information about which AI models are accessible for use in workflow agent steps.
The model information includes both the provider identifier and the list of model IDs that can be referenced in workflow definitions.
type SchemaOutput ¶
type SchemaOutput struct {
// Schema contains the JSON Schema definition for Lacquer workflow files.
// This schema can be used to validate workflow YAML/JSON files and
// provide autocompletion and validation in editors.
Schema json.RawMessage `json:"schema"`
// Expressions lists all available expression types and their syntax definitions.
// This includes template expressions, conditional expressions, and data
// transformation expressions that can be used within workflow definitions.
Expressions []expression.ExpressionDef `json:"expressions"`
// Functions contains all built-in functions available for use in expressions.
// These functions provide data manipulation, string processing, mathematical
// operations, and other utility functions within workflow expressions.
Functions []*expression.FunctionDefinition `json:"functions"`
// ModelProviders lists all available AI model providers and their supported models.
// This information is essential for knowing which AI models can be referenced
// in workflow agent configurations.
ModelProviders []Model `json:"model_providers"`
}
SchemaOutput represents the complete schema information for Lacquer workflows. This structure contains all the metadata needed to understand and validate Lacquer workflow definitions, including the JSON schema, expression syntax, available functions, and supported AI model providers.
func GetSchema ¶
func GetSchema() (*SchemaOutput, error)
GetSchema retrieves the complete schema information for Lacquer workflows. This function compiles and returns all metadata necessary for understanding and working with Lacquer workflow definitions.
The function performs the following operations:
- Generates the base JSON schema for workflow validation
- Initializes all supported AI model providers
- Queries each provider for their available models
- Compiles expression and function definitions
- Returns the complete schema metadata
This is the primary entry point for accessing Lacquer's schema information and is typically used by development tools, validators, and documentation generators.
Returns:
- *SchemaOutput: Complete schema information including JSON schema, expressions, functions, and available AI models
- error: Any error that occurred during schema generation or provider initialization
Errors can occur due to:
- Schema generation failures
- AI provider initialization issues
- Network failures when querying model availability
- Provider authentication or configuration problems
Example:
schema, err := GetSchema()
if err != nil {
return fmt.Errorf("failed to get schema: %w", err)
}
// Use the schema for validation
var workflowSchema map[string]interface{}
if err := json.Unmarshal(schema.Schema, &workflowSchema); err != nil {
return fmt.Errorf("failed to parse schema: %w", err)
}
// Check available models
fmt.Printf("Available models: %d providers\n", len(schema.ModelProviders))
for _, provider := range schema.ModelProviders {
fmt.Printf(" %s: %d models\n", provider.Provider, len(provider.Models))
}
// Access expression definitions
fmt.Printf("Supported expressions: %d types\n", len(schema.Expressions))
fmt.Printf("Built-in functions: %d functions\n", len(schema.Functions))
Source Files