iatoolkit 0.66.7__tar.gz → 0.69.0__tar.gz

iatoolkit-0.69.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 Fernando Libedinsky
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in
+all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+THE SOFTWARE.

{iatoolkit-0.66.7 → iatoolkit-0.69.0}/PKG-INFO

@@ -1,11 +1,12 @@
 Metadata-Version: 2.4
 Name: iatoolkit
-Version: 0.66.7
+Version: 0.69.0
 Summary: IAToolkit
 Author: Fernando Libedinsky
 License-Expression: MIT
 Requires-Python: >=3.12
 Description-Content-Type: text/markdown
+License-File: LICENSE
 Requires-Dist: aiohappyeyeballs==2.4.4
 Requires-Dist: aiohttp==3.11.9
 Requires-Dist: aiosignal==1.3.1
@@ -206,47 +207,64 @@ Requires-Dist: wrapt==1.17.0
 Requires-Dist: yarl==1.18.3
 Requires-Dist: zipp==3.21.0
 Requires-Dist: zstandard==0.23.0
+Dynamic: license-file
 
 
 <div align="center">
-  <h1>IAToolkit</h1>
+<h1>
+    IAToolkit
+</h1>
+
   <p><strong>The Open-Source Framework for Building AI Chatbots on Your Private Data.</strong></p>
+<h4>
+  <a href="https://www.iatoolkit.com" target="_blank" style="text-decoration: none; color: inherit;">
+    www.iatoolkit.com
+  </a>
+</h4>
 </div>
 
-IAToolkit is a comprehensive, open-source framework designed for building enterprise-grade 
-AI chatbots and conversational applications. 
+IAToolkit is a comprehensive, Python open-source framework designed for building enterprise-grade 
+AI chatbots and conversational applications. It bridges the gap between the power of 
+Large Language Models (LLMs) and the valuable, 
+private data locked within your organization's databases and documents.
+
 With IAToolkit, you can build production-ready, context-aware chatbots and agents that 
 can query relational databases, perform semantic searches on documents, 
 and connect to your internal APIs in minutes.
 
-IAToolkit bridges the gap between powerful LLMs and your company's data. 
+Create secure, branded chat interfaces that can reason over your data, answer questions, and execute custom business logic, 
+all powered by leading models from OpenAI, Google Gemini, and more.
 
 
 ## 🚀 Key Features
 
-*   **🔗 Unified Data Connection**:
-    *   **Natural Language to SQL**: Let your chatbot query relational databases (PostgreSQL, MySQL, SQLite) using everyday language.
-    *   **Semantic Document Search**: Automatically chunk, embed, and search across your private documents (PDFs, Word, etc.) to provide contextually accurate answers.
+* **🔗 Unified Data Connection**
+    * **Natural Language to SQL**: Let your chatbot query relational databases (PostgreSQL, MySQL, SQLite) using everyday language.
+    * **Semantic Document Search**: Automatically chunk, embed, and search across your private documents (PDFs, Word, etc.) to provide contextually accurate answers.
 
-*   **🏢 Enterprise-Ready Multi-Tenancy**:
-    *   Deploy isolated "Company" modules, each with its own data, tools, and context. Perfect for SaaS products or internal departmental agents.
+* **🏢 Enterprise-Ready Multi-Tenancy**
+    * Deploy isolated "Company" modules, each with its own data, tools, and context. 
+    * Perfect for SaaS products or internal departmental agents.
 
-*   **🧠 LLM Agnostic**:
-    *   Switch between **OpenAI (GPT-*)** and **Google (Gemini-*)** with a single line change in your configuration. No code refactoring needed.
+* **🎨 Fully Brandable UI**
+    * Customize the look and feel for each "Company" with its own logos, colors, and even language settings (i18n).
+    * Provides a white-labeled experience for your users.
 
-*   **🛠️ Developer-First Experience**:
-    *   Built with a clean, **Dependency Injection** architecture.
-    *   High-quality code base with **90%+ test coverage**.
-    *   Powerful Flask-based **CLI** for database setup, API key generation, and more.
+* **🧠 LLM Agnostic**
+    * Switch between **OpenAI (GPT-*)** and **Google (Gemini-*)** with a single line change in your configuration. 
+    * No code refactoring needed.
 
-*   **🔒 Security & Observability Built-In**:
-    *   Comes with JWT-based authentication, user management, and secure session handling out of the box.
-    *   Full traceability with detailed logging of all queries, function calls, token usage, and costs.
+* **🛠️ Developer-First Experience**
+    * Built with a clean **Dependency Injection** architecture.  
+    * High-quality code base with **90%+ test coverage**.
 
-## ⚡ Quick Start: Create a Custom Tool in 30 Seconds
+* **🔒 Security & Observability Built-In**
+    * Comes with integrated user authentication, API keys, and secure session handling out of the box.  
+    * Full traceability with detailed logging of all queries, function calls, token usage, and costs.
+## ⚡ Quick Start: Try our 'hello world' example
 
-See how easy it is to give your AI a new skill. Just define a method inside your Company class and describe it. 
-IAToolkit handles the rest.
+Ready to see it in action? Our Quickstart Guide will walk you through downloading, configuring, and launching your first AI assistant in just a few minutes. 
+It's the best way to experience the toolkit's capabilities firsthand.
 
 ## 🤝 Contributing
 

{iatoolkit-0.66.7 → iatoolkit-0.69.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "iatoolkit"
-version = "0.66.7"
+version = "0.69.0"
 requires-python = ">=3.12"
 description = "IAToolkit"
 readme = "readme.md"

iatoolkit-0.69.0/readme.md

@@ -0,0 +1,65 @@
+
+<div align="center">
+<h1>
+    IAToolkit
+</h1>
+
+  <p><strong>The Open-Source Framework for Building AI Chatbots on Your Private Data.</strong></p>
+<h4>
+  <a href="https://www.iatoolkit.com" target="_blank" style="text-decoration: none; color: inherit;">
+    www.iatoolkit.com
+  </a>
+</h4>
+</div>
+
+IAToolkit is a comprehensive, Python open-source framework designed for building enterprise-grade 
+AI chatbots and conversational applications. It bridges the gap between the power of 
+Large Language Models (LLMs) and the valuable, 
+private data locked within your organization's databases and documents.
+
+With IAToolkit, you can build production-ready, context-aware chatbots and agents that 
+can query relational databases, perform semantic searches on documents, 
+and connect to your internal APIs in minutes.
+
+Create secure, branded chat interfaces that can reason over your data, answer questions, and execute custom business logic, 
+all powered by leading models from OpenAI, Google Gemini, and more.
+
+
+## 🚀 Key Features
+
+* **🔗 Unified Data Connection**
+    * **Natural Language to SQL**: Let your chatbot query relational databases (PostgreSQL, MySQL, SQLite) using everyday language.
+    * **Semantic Document Search**: Automatically chunk, embed, and search across your private documents (PDFs, Word, etc.) to provide contextually accurate answers.
+
+* **🏢 Enterprise-Ready Multi-Tenancy**
+    * Deploy isolated "Company" modules, each with its own data, tools, and context. 
+    * Perfect for SaaS products or internal departmental agents.
+
+* **🎨 Fully Brandable UI**
+    * Customize the look and feel for each "Company" with its own logos, colors, and even language settings (i18n).
+    * Provides a white-labeled experience for your users.
+
+* **🧠 LLM Agnostic**
+    * Switch between **OpenAI (GPT-*)** and **Google (Gemini-*)** with a single line change in your configuration. 
+    * No code refactoring needed.
+
+* **🛠️ Developer-First Experience**
+    * Built with a clean **Dependency Injection** architecture.  
+    * High-quality code base with **90%+ test coverage**.
+
+* **🔒 Security & Observability Built-In**
+    * Comes with integrated user authentication, API keys, and secure session handling out of the box.  
+    * Full traceability with detailed logging of all queries, function calls, token usage, and costs.
+## ⚡ Quick Start: Try our 'hello world' example
+
+Ready to see it in action? Our Quickstart Guide will walk you through downloading, configuring, and launching your first AI assistant in just a few minutes. 
+It's the best way to experience the toolkit's capabilities firsthand.
+
+## 🤝 Contributing
+
+We welcome contributions! Whether it's adding a new feature, improving documentation, or fixing a bug, 
+please feel free to open a pull request.
+
+## 📄 License
+
+IAToolkit is open-source and licensed under the [MIT License](LICENSE).

{iatoolkit-0.66.7 → iatoolkit-0.69.0}/src/iatoolkit/__init__.py

@@ -10,7 +10,6 @@ from .iatoolkit import IAToolkit, current_iatoolkit, create_app
 # for registering the client companies
 from .company_registry import register_company
 from .base_company import BaseCompany
-from iatoolkit.repositories.database_manager import DatabaseManager
 
 # --- Services ---
 from iatoolkit.services.query_service import QueryService
@@ -27,7 +26,6 @@ __all__ = [
     'current_iatoolkit',
     'register_company',
     'BaseCompany',
-    'DatabaseManager',
     'QueryService',
     'SqlService',
     'ExcelService',

{iatoolkit-0.66.7 → iatoolkit-0.69.0}/src/iatoolkit/base_company.py

@@ -7,7 +7,6 @@
 from abc import ABC, abstractmethod
 from iatoolkit.repositories.profile_repo import ProfileRepo
 from iatoolkit.repositories.llm_query_repo import LLMQueryRepo
-
 from iatoolkit.services.prompt_manager_service import PromptService
 from iatoolkit.repositories.models import Company, Function, PromptCategory
 from iatoolkit import IAToolkit
@@ -22,22 +21,14 @@ class BaseCompany(ABC):
         self.prompt_service: PromptService = injector.get(PromptService)
         self.company: Company | None = None
 
-    def _load_company_by_short_name(self, short_name: str) -> Company:
-        self.company = self.profile_repo.get_company_by_short_name(short_name)
-        return self.company
-
     def _create_company(self,
                         short_name: str,
                         name: str,
                         parameters: dict | None = None,
-                        branding: dict | None = None,
-                        onboarding_cards: dict | None = None,
                         ) -> Company:
         company_obj = Company(short_name=short_name,
                               name=name,
-                              parameters=parameters,
-                              branding=branding,
-                              onboarding_cards=onboarding_cards)
+                              parameters=parameters)
         self.company = self.profile_repo.create_company(company_obj)
         return self.company
 
@@ -77,17 +68,6 @@ class BaseCompany(ABC):
             **kwargs
         )
 
-
-    @abstractmethod
-    # initialize all the database tables  needed
-    def register_company(self):
-        raise NotImplementedError("La subclase debe implementar el método create_company()")
-
-    @abstractmethod
-    # get context specific for this company
-    def get_company_context(self, **kwargs) -> str:
-        raise NotImplementedError("La subclase debe implementar el método get_company_context()")
-
     @abstractmethod
     # get context specific for this company
     def get_user_info(self, user_identifier: str) -> dict:
@@ -98,11 +78,6 @@ class BaseCompany(ABC):
     def handle_request(self, tag: str, params: dict) -> dict:
         raise NotImplementedError("La subclase debe implementar el método handle_request()")
 
-    @abstractmethod
-    # get context specific for the query
-    def start_execution(self):
-        raise NotImplementedError("La subclase debe implementar el método start_execution()")
-
     @abstractmethod
     # get context specific for the query
     def get_metadata_from_filename(self, filename: str) -> dict:

{iatoolkit-0.66.7 → iatoolkit-0.69.0}/src/iatoolkit/company_registry.py

@@ -31,10 +31,9 @@ class CompanyRegistry:
 
                     # save the created instance in the registry
                     self._company_instances[company_key] = company_instance
-                    logging.info(f"company '{company_key}' instantiated")
 
                 except Exception as e:
-                    logging.error(f"Error instanciando empresa {company_key}: {e}")
+                    logging.error(f"Error while creating company instance for {company_key}: {e}")
                     logging.exception(e)
                     raise
 

{iatoolkit-0.66.7 → iatoolkit-0.69.0}/src/iatoolkit/iatoolkit.py

@@ -19,7 +19,7 @@ from werkzeug.middleware.proxy_fix import ProxyFix
 from injector import Binder, Injector, singleton
 from importlib.metadata import version as _pkg_version, PackageNotFoundError
 
-IATOOLKIT_VERSION = "0.66.7"
+IATOOLKIT_VERSION = "0.69.0"
 
 # global variable for the unique instance of IAToolkit
 _iatoolkit_instance: Optional['IAToolkit'] = None
@@ -293,7 +293,6 @@ class IAToolkit:
         from iatoolkit.repositories.document_repo import DocumentRepo
         from iatoolkit.repositories.profile_repo import ProfileRepo
         from iatoolkit.repositories.llm_query_repo import LLMQueryRepo
-
         from iatoolkit.repositories.vs_repo import VSRepo
         from iatoolkit.repositories.tasks_repo import TaskRepo
 
@@ -318,6 +317,7 @@ class IAToolkit:
         from iatoolkit.services.branding_service import BrandingService
         from iatoolkit.services.i18n_service import I18nService
         from iatoolkit.services.language_service import LanguageService
+        from iatoolkit.services.configuration_service import ConfigurationService
 
         binder.bind(QueryService, to=QueryService)
         binder.bind(TaskService, to=TaskService)
@@ -333,6 +333,7 @@ class IAToolkit:
         binder.bind(BrandingService, to=BrandingService)
         binder.bind(I18nService, to=I18nService)
         binder.bind(LanguageService, to=LanguageService)
+        binder.bind(ConfigurationService, to=ConfigurationService)
 
     def _bind_infrastructure(self, binder: Binder):
         from iatoolkit.infra.llm_client import llmClient
@@ -359,9 +360,9 @@ class IAToolkit:
         # instantiate all the registered companies
         get_company_registry().instantiate_companies(self._injector)
 
-        # use the dispatcher to start the execution of every company
+        # use the dispatcher to load the company config.yaml file and prepare the execution
         dispatcher = self._injector.get(Dispatcher)
-        dispatcher.start_execution()
+        dispatcher.load_company_configs()
 
     def _setup_cli_commands(self):
         from iatoolkit.cli_commands import register_core_commands

{iatoolkit-0.66.7 → iatoolkit-0.69.0}/src/iatoolkit/repositories/database_manager.py

@@ -70,6 +70,11 @@ class DatabaseManager:
     def remove_session(self):
         self.scoped_session.remove()
 
+    def get_all_table_names(self) -> list[str]:
+        # Returns a list of all table names in the database
+        inspector = inspect(self._engine)
+        return inspector.get_table_names()
+
     def get_table_schema(self,
                          table_name: str,
                          schema_name: str | None = None,

{iatoolkit-0.66.7 → iatoolkit-0.69.0}/src/iatoolkit/repositories/models.py

@@ -53,17 +53,12 @@ class Company(Base):
     id = Column(Integer, primary_key=True)
     short_name = Column(String(20), nullable=False, unique=True, index=True)
     name = Column(String(256), nullable=False)
-    default_language = Column(String(5), nullable=False, default='es')
 
     # encrypted api-key
     openai_api_key = Column(String, nullable=True)
     gemini_api_key = Column(String, nullable=True)
-
-    branding = Column(JSON, nullable=True)
-    onboarding_cards = Column(JSON, nullable=True)
     parameters = Column(JSON, nullable=True)
     created_at = Column(DateTime, default=datetime.now)
-    allow_jwt = Column(Boolean, default=True, nullable=True)
 
     documents = relationship("Document",
                              back_populates="company",

{iatoolkit-0.66.7 → iatoolkit-0.69.0}/src/iatoolkit/repositories/profile_repo.py

@@ -74,10 +74,6 @@ class ProfileRepo:
         if company:
             if company.parameters != new_company.parameters:
                 company.parameters = new_company.parameters
-            if company.branding != new_company.branding:
-                company.branding = new_company.branding
-            if company.onboarding_cards != new_company.onboarding_cards:
-                company.onboarding_cards = new_company.onboarding_cards
         else:
             # Si la compañía no existe, la añade a la sesión.
             self.session.add(new_company)

{iatoolkit-0.66.7 → iatoolkit-0.69.0}/src/iatoolkit/services/branding_service.py

@@ -4,6 +4,7 @@
 # IAToolkit is open source software.
 
 from iatoolkit.repositories.models import Company
+from iatoolkit.services.configuration_service import ConfigurationService
 from injector import inject
 
 
@@ -12,7 +13,8 @@ class BrandingService:
     Branding configuration for IAToolkit
     """
     @inject
-    def __init__(self):
+    def __init__(self, config_service: ConfigurationService):
+        self.config_service = config_service
         """
         Define los estilos de branding por defecto para la aplicación.
         """
@@ -67,15 +69,15 @@ class BrandingService:
             "send_button_color": "#212529"          # Gris oscuro/casi negro por defecto
         }
 
-    def get_company_branding(self, company: Company | None) -> dict:
+    def get_company_branding(self, company_short_name: str) -> dict:
         """
         Retorna los estilos de branding finales para una compañía,
         fusionando los valores por defecto con los personalizados.
         """
         final_branding_values = self._default_branding.copy()
+        branding_data = self.config_service.get_configuration(company_short_name, 'branding')
+        final_branding_values.update(branding_data)
 
-        if company and company.branding:
-            final_branding_values.update(company.branding)
 
         # Función para convertir HEX a RGB
         def hex_to_rgb(hex_color):
@@ -137,8 +139,11 @@ class BrandingService:
             }}
         """
 
+        # get the company name from configuration for the branding render
+        company_name = self.config_service.get_configuration(company_short_name, 'name')
+
         return {
-            "name": company.name if company else "IAToolkit",
+            "name": company_name,
             "primary_text_style": primary_text_style,
             "secondary_text_style": secondary_text_style,
             "tertiary_text_style": tertiary_text_style,

iatoolkit-0.69.0/src/iatoolkit/services/company_context_service.py

@@ -0,0 +1,145 @@
+# Copyright (c) 2024 Fernando Libedinsky
+# Product: IAToolkit
+#
+# IAToolkit is open source software.
+
+from iatoolkit.common.util import Utility
+from iatoolkit.services.configuration_service import ConfigurationService
+from iatoolkit.services.sql_service import SqlService
+from iatoolkit.common.exceptions import IAToolkitException
+import logging
+from injector import inject
+import os
+
+
+class CompanyContextService:
+    """
+    Responsible for building the complete context string for a given company
+    to be sent to the Language Model.
+    """
+
+    @inject
+    def __init__(self,
+                 sql_service: SqlService,
+                 utility: Utility,
+                 config_service: ConfigurationService):
+        self.sql_service = sql_service
+        self.utility = utility
+        self.config_service = config_service
+
+    def get_company_context(self, company_short_name: str) -> str:
+        """
+        Builds the full context by aggregating three sources:
+        1. Static context files (Markdown).
+        2. Static schema files (YAML for APIs, etc.).
+        3. Dynamic SQL database schema from the live connection.
+        """
+        context_parts = []
+
+        # 1. Context from Markdown (context/*.md) and yaml (schema/*.yaml) files
+        try:
+            md_context = self._get_static_file_context(company_short_name)
+            if md_context:
+                context_parts.append(md_context)
+        except Exception as e:
+            logging.warning(f"Could not load Markdown context for '{company_short_name}': {e}")
+
+        # 2. Context from company-specific Python logic (SQL schemas)
+        try:
+            sql_context = self._get_sql_schema_context(company_short_name)
+            if sql_context:
+                context_parts.append(sql_context)
+        except Exception as e:
+            logging.warning(f"Could not generate SQL context for '{company_short_name}': {e}")
+
+        # Join all parts with a clear separator
+        return "\n\n---\n\n".join(context_parts)
+
+    def _get_static_file_context(self, company_short_name: str) -> str:
+        # Get context from .md and .yaml schema files.
+        static_context = ''
+
+        # Part 1: Markdown context files
+        context_dir = f'companies/{company_short_name}/context'
+        if os.path.exists(context_dir):
+            context_files = self.utility.get_files_by_extension(context_dir, '.md', return_extension=True)
+            for file in context_files:
+                filepath = os.path.join(context_dir, file)
+                static_context += self.utility.load_markdown_context(filepath)
+
+        # Part 2: YAML schema files
+        schema_dir = f'companies/{company_short_name}/schema'
+        if os.path.exists(schema_dir):
+            schema_files = self.utility.get_files_by_extension(schema_dir, '.yaml', return_extension=True)
+            for file in schema_files:
+                schema_name = file.split('.')[0]  # Use full filename as entity name
+                filepath = os.path.join(schema_dir, file)
+                static_context += self.utility.generate_context_for_schema(schema_name, filepath)
+
+        return static_context
+
+    def _get_sql_schema_context(self, company_short_name: str) -> str:
+        """
+        Generates the SQL schema context by inspecting live database connections
+        based on the flexible company.yaml configuration.
+        It supports including all tables and providing specific overrides for a subset of them.
+        """
+        data_sources_config = self.config_service.get_configuration(company_short_name, 'data_sources')
+        if not data_sources_config or not data_sources_config.get('sql'):
+            return ''
+
+        sql_context = ''
+        for source in data_sources_config.get('sql', []):
+            db_name = source.get('database')
+            if not db_name:
+                continue
+
+            try:
+                db_manager = self.sql_service.get_database_manager(db_name)
+            except IAToolkitException as e:
+                logging.warning(f"Could not get DB manager for '{db_name}': {e}")
+                continue
+
+            db_description = source.get('description', '')
+            sql_context += f"{db_description}\n" if db_description else ""
+
+            # 1. get the list of tables to process.
+            tables_to_process = []
+            if source.get('include_all_tables', False):
+                all_tables = db_manager.get_all_table_names()
+                tables_to_exclude = set(source.get('exclude_tables', []))
+                tables_to_process = [t for t in all_tables if t not in tables_to_exclude]
+            elif 'tables' in source:
+                # if not include_all_tables, use the list of tables explicitly specified in the map.
+                tables_to_process = list(source['tables'].keys())
+
+            # 2. get the global list of columns to exclude.
+            global_exclude_columns = source.get('exclude_columns', [])
+
+            # 3. get the overrides for specific tables.
+            table_overrides = source.get('tables', {})
+
+            # 3. iterate over the tables.
+            for table_name in tables_to_process:
+                try:
+                    # 4. get the table specific configuration.
+                    table_config = table_overrides.get(table_name, {})
+
+                    # 5. define the schema name, using the override if it exists.
+                    schema_name = table_config.get('schema_name', table_name)
+
+                    # 6. define the list of columns to exclude, (local vs. global).
+                    local_exclude_columns = table_config.get('exclude_columns')
+                    final_exclude_columns = local_exclude_columns if local_exclude_columns is not None else global_exclude_columns
+
+                    # 7. get the table schema definition.
+                    table_definition = db_manager.get_table_schema(
+                        table_name=table_name,
+                        schema_name=schema_name,
+                        exclude_columns=final_exclude_columns
+                    )
+                    sql_context += table_definition
+                except (KeyError, RuntimeError) as e:
+                    logging.warning(f"Could not generate schema for table '{table_name}': {e}")
+
+        return sql_context