Skip to content

api_reference.md

Latest commit

 

History

History
250 lines (161 loc) · 4.33 KB

api_reference.md

File metadata and controls

250 lines (161 loc) · 4.33 KB
Raw

API Reference

Complete API documentation for the GENXAIS Framework.

Core Framework

GENXAISFramework

Main framework class that provides access to all GENXAIS features.

from genxais_sdk import GENXAISFramework

framework = GENXAISFramework(config_path: Optional[str] = None)

Methods

set_mode(mode: str) -> bool

Set the current development mode.

Parameters:

  • mode: One of "VAN", "PLAN", "CREATE", "IMPLEMENT", "REFLECT", "ARCHIVE"

Returns:

  • bool: Success status
get_mode() -> str

Get the current development mode.

Returns:

  • str: Current mode

RAG System

RAGStorageInitializer

Initialize and manage RAG storage components.

from rag_system.init_storage import RAGStorageInitializer

initializer = RAGStorageInitializer(mongodb_uri: Optional[str] = None)

Methods

init_filesystem() -> Dict[str, Any]

Initialize filesystem structure.

Returns:

  • Dictionary with initialization status
init_mongodb() -> Dict[str, Any]

Initialize MongoDB collections and indexes.

Returns:

  • Dictionary with initialization status
initialize_all() -> Dict[str, Any]

Initialize all RAG components.

Returns:

  • Dictionary with complete initialization status

Error Handling

SDKErrorHandler

Manage and recover from errors.

from error_handling import SDKErrorHandler

handler = SDKErrorHandler()

Methods

handle_error(error_type: str, error_details: Dict[str, Any], context: str = "") -> Dict[str, Any]

Handle specific error types with recovery strategies.

Parameters:

  • error_type: Type of error
  • error_details: Error details
  • context: Error context

Returns:

  • Dictionary with error handling results
safe_execute(func: Callable, *args, **kwargs) -> Dict[str, Any]

Safely execute a function with error handling.

Parameters:

  • func: Function to execute
  • *args: Positional arguments
  • **kwargs: Keyword arguments

Returns:

  • Dictionary with execution results

APM Framework

APMCycleManager

Manage APM development cycles.

from apm_framework import APMCycleManager

manager = APMCycleManager()

Methods

start_cycle(mode: str) -> Dict[str, Any]

Start a new APM cycle.

Parameters:

  • mode: Development mode

Returns:

  • Dictionary with cycle status
end_cycle() -> Dict[str, Any]

End current APM cycle.

Returns:

  • Dictionary with cycle results

Agent System

BaseAgent

Base class for all agents.

from agents import BaseAgent

class CustomAgent(BaseAgent):
    def __init__(self):
        super().__init__()

Methods

validate_action(action: str, context: Dict[str, Any]) -> bool

Validate if action is allowed in current mode.

Parameters:

  • action: Proposed action
  • context: Action context

Returns:

  • bool: Action validity

Memory Bank

MemoryManager

Manage development context and history.

from memory_bank import MemoryManager

memory = MemoryManager()

Methods

store_context(context: Dict[str, Any]) -> str

Store development context.

Parameters:

  • context: Context data

Returns:

  • str: Context ID
retrieve_context(context_id: str) -> Dict[str, Any]

Retrieve stored context.

Parameters:

  • context_id: Context identifier

Returns:

  • Dictionary with context data

Utility Functions

Token Optimization

from genxais_sdk.utils import optimize_tokens

optimized = optimize_tokens(text: str, max_tokens: int) -> str

Pipeline Management

from genxais_sdk.utils import create_pipeline

pipeline = create_pipeline(steps: List[Dict[str, Any]]) -> Pipeline

Event System

EventEmitter

from genxais_sdk.events import EventEmitter

emitter = EventEmitter()
emitter.on("event_name", callback_function)
emitter.emit("event_name", event_data)

Configuration

ConfigManager

from genxais_sdk.config import ConfigManager

config = ConfigManager()
config.load_config("config.json")
config.get_value("key")
config.set_value("key", "value")

Constants

from genxais_sdk.constants import *

MODES = ["VAN", "PLAN", "CREATE", "IMPLEMENT", "REFLECT", "ARCHIVE"]
DEFAULT_TIMEOUT = 60
MAX_RETRIES = 3

Type Definitions

from genxais_sdk.types import *

Pipeline = List[Dict[str, Any]]
Context = Dict[str, Any]
ActionResult = Dict[str, Any]