memra 0.2.0__tar.gz → 0.2.1__tar.gz

memra-0.2.1/CHANGELOG.md

@@ -0,0 +1,44 @@
+# Changelog
+
+All notable changes to the Memra SDK will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [0.2.0] - 2024-01-17
+
+### Added
+- **MCP (Model Context Protocol) Integration**: Execute operations on local infrastructure while leveraging cloud AI processing
+  - New `mcp_bridge_server.py` for local resource bridging
+  - HMAC authentication for secure cloud-to-local communication
+  - Support for `hosted_by: "mcp"` in agent tool configurations
+  - PostgreSQL integration via MCP bridge
+  - Tool-level configuration support in execution engine
+  - New MCP tools: `PostgresInsert`, `DataValidator`
+
+### Enhanced
+- **Execution Engine**: Updated to support tool-level configuration and MCP routing
+- **Tool Registry Client**: Enhanced API client with better error handling and MCP support
+- **Agent Configuration**: Added support for tool-specific configuration alongside agent-level config
+
+### Examples
+- `examples/accounts_payable_mcp.py` - Complete invoice processing with MCP database integration
+- `test_mcp_success.py` - Simple MCP integration test
+
+### Documentation
+- `docs/mcp_integration.md` - Comprehensive MCP integration guide
+- Updated README with MCP overview and quick start
+
+### Dependencies
+- Added `aiohttp>=3.8.0` for MCP bridge server
+- Added `aiohttp-cors>=0.7.0` for CORS support
+- Added `psycopg2-binary>=2.9.0` for PostgreSQL integration
+
+## [0.1.0] - 2024-01-01
+
+### Added
+- Initial release of Memra SDK
+- Core agent and department framework
+- API client for Memra cloud services
+- Basic tool registry and execution engine
+- Examples for accounts payable and propane delivery workflows 

memra-0.2.1/MANIFEST.in

@@ -0,0 +1,29 @@
+# Include only the client SDK files
+include README.md
+include LICENSE
+include CHANGELOG.md
+recursive-include memra *.py
+
+# Explicitly exclude server-only files and directories
+exclude app.py
+exclude server_tool_registry.py
+exclude mcp_bridge_server.py
+exclude config.py
+exclude fly.toml
+exclude Dockerfile
+exclude Procfile
+exclude docker-compose.yml
+exclude requirements.txt
+recursive-exclude logic *
+recursive-exclude scripts *
+recursive-exclude docs *
+recursive-exclude examples *
+recursive-exclude tests *
+recursive-exclude temp_processing *
+recursive-exclude invoices *
+recursive-exclude local *
+recursive-exclude dist *
+recursive-exclude __pycache__ *
+recursive-exclude *.egg-info *
+exclude .DS_Store
+exclude .dockerignore 

{memra-0.2.0/memra.egg-info → memra-0.2.1}/PKG-INFO

@@ -1,10 +1,10 @@
-Metadata-Version: 2.4
+Metadata-Version: 2.1
 Name: memra
-Version: 0.2.0
-Summary: Declarative framework for enterprise workflows with MCP integration
+Version: 0.2.1
+Summary: Declarative framework for enterprise workflows with MCP integration - Client SDK
 Home-page: https://github.com/memra/memra-sdk
 Author: Memra
-Author-email: Memra <info@memra.co>
+Author-email: Memra <support@memra.com>
 License: MIT
 Project-URL: Homepage, https://memra.co
 Project-URL: Repository, https://github.com/memra-platform/memra-sdk
@@ -25,16 +25,13 @@ Requires-Dist: httpx>=0.24.0
 Requires-Dist: typing-extensions>=4.0.0
 Requires-Dist: aiohttp>=3.8.0
 Requires-Dist: aiohttp-cors>=0.7.0
-Requires-Dist: psycopg2-binary>=2.9.0
 Provides-Extra: dev
 Requires-Dist: pytest>=6.0; extra == "dev"
 Requires-Dist: pytest-asyncio; extra == "dev"
 Requires-Dist: black; extra == "dev"
 Requires-Dist: flake8; extra == "dev"
-Dynamic: author
-Dynamic: home-page
-Dynamic: license-file
-Dynamic: requires-python
+Provides-Extra: mcp
+Requires-Dist: psycopg2-binary>=2.9.0; extra == "mcp"
 
 # Memra SDK
 
@@ -131,31 +128,3 @@ This project is licensed under the MIT License - see the [LICENSE](LICENSE) file
 ├── local/dependencies/     # Database setup & schemas
 └── docker-compose.yml      # Database setup
 ```
-
-## ✨ New: MCP Integration
-
-Memra now supports **Model Context Protocol (MCP)** integration, allowing you to execute operations on your local infrastructure while leveraging Memra's cloud-based AI processing.
-
-**Key Benefits:**
-- 🔒 **Keep sensitive data local** - Your databases stay on your infrastructure
-- ⚡ **Hybrid processing** - AI processing in the cloud, data operations locally  
-- 🔐 **Secure communication** - HMAC-authenticated requests between cloud and local
-- 🛠️ **Easy setup** - Simple bridge server connects your local resources
-
-**Quick Example:**
-```python
-# Agent that uses local database via MCP
-agent = Agent(
-    role="Data Writer",
-    tools=[{
-        "name": "PostgresInsert",
-        "hosted_by": "mcp",  # Routes to your local infrastructure
-        "config": {
-            "bridge_url": "http://localhost:8081",
-            "bridge_secret": "your-secret"
-        }
-    }]
-)
-```
-
-📖 **[Complete MCP Integration Guide →](docs/mcp_integration.md)**

{memra-0.2.0 → memra-0.2.1}/README.md

@@ -93,31 +93,3 @@ This project is licensed under the MIT License - see the [LICENSE](LICENSE) file
 ├── local/dependencies/     # Database setup & schemas
 └── docker-compose.yml      # Database setup
 ```
-
-## ✨ New: MCP Integration
-
-Memra now supports **Model Context Protocol (MCP)** integration, allowing you to execute operations on your local infrastructure while leveraging Memra's cloud-based AI processing.
-
-**Key Benefits:**
-- 🔒 **Keep sensitive data local** - Your databases stay on your infrastructure
-- ⚡ **Hybrid processing** - AI processing in the cloud, data operations locally  
-- 🔐 **Secure communication** - HMAC-authenticated requests between cloud and local
-- 🛠️ **Easy setup** - Simple bridge server connects your local resources
-
-**Quick Example:**
-```python
-# Agent that uses local database via MCP
-agent = Agent(
-    role="Data Writer",
-    tools=[{
-        "name": "PostgresInsert",
-        "hosted_by": "mcp",  # Routes to your local infrastructure
-        "config": {
-            "bridge_url": "http://localhost:8081",
-            "bridge_secret": "your-secret"
-        }
-    }]
-)
-```
-
-📖 **[Complete MCP Integration Guide →](docs/mcp_integration.md)**

memra-0.2.1/memra/__init__.py

@@ -0,0 +1,27 @@
+"""
+Memra SDK - Declarative AI Workflows
+
+A framework for building AI-powered business workflows using a declarative approach.
+Think of it as "Kubernetes for business logic" where agents are the pods and 
+departments are the deployments.
+"""
+
+__version__ = "0.2.1"
+
+# Core imports
+from .models import Agent, Department, Tool
+from .execution import ExecutionEngine
+
+# Make key classes available at package level
+__all__ = [
+    "Agent",
+    "Department", 
+    "Tool",
+    "ExecutionEngine",
+    "__version__"
+]
+
+# Optional: Add version check for compatibility
+import sys
+if sys.version_info < (3, 8):
+    raise RuntimeError("Memra requires Python 3.8 or higher") 

{memra-0.2.0/memra-sdk-package → memra-0.2.1}/memra/execution.py

@@ -2,7 +2,7 @@ import time
 import logging
 from typing import Dict, Any, List, Optional
 from .models import Department, Agent, DepartmentResult, ExecutionTrace, DepartmentAudit
-from .tool_registry_client import ToolRegistryClient
+from .tool_registry import ToolRegistry
 
 logger = logging.getLogger(__name__)
 
@@ -10,7 +10,7 @@ class ExecutionEngine:
     """Engine that executes department workflows by coordinating agents and tools"""
     
     def __init__(self):
-        self.tool_registry = ToolRegistryClient()
+        self.tool_registry = ToolRegistry()
         self.last_execution_audit: Optional[DepartmentAudit] = None
     
     def execute_department(self, department: Department, input_data: Dict[str, Any]) -> DepartmentResult:

memra-0.2.1/memra/tool_registry.py

@@ -0,0 +1,70 @@
+import importlib
+import logging
+import sys
+import os
+from typing import Dict, Any, List, Optional, Callable
+from pathlib import Path
+
+logger = logging.getLogger(__name__)
+
+class ToolRegistry:
+    """Registry for managing and executing tools via API calls only"""
+    
+    def __init__(self):
+        self.tools: Dict[str, Dict[str, Any]] = {}
+        self._register_known_tools()
+    
+    def _register_known_tools(self):
+        """Register known tools with their metadata (no actual implementations)"""
+        # Server-hosted tools (executed via Memra API)
+        server_tools = [
+            ("DatabaseQueryTool", "Query database schemas and data"),
+            ("PDFProcessor", "Process PDF files and extract content"),
+            ("OCRTool", "Perform OCR on images and documents"),
+            ("InvoiceExtractionWorkflow", "Extract structured data from invoices"),
+            ("FileReader", "Read files from the filesystem"),
+        ]
+        
+        for tool_name, description in server_tools:
+            self.register_tool(tool_name, None, "memra", description)
+        
+        # MCP-hosted tools (executed via MCP bridge)
+        mcp_tools = [
+            ("DataValidator", "Validate data against schemas"),
+            ("PostgresInsert", "Insert data into PostgreSQL database"),
+        ]
+        
+        for tool_name, description in mcp_tools:
+            self.register_tool(tool_name, None, "mcp", description)
+        
+        logger.info(f"Registered {len(self.tools)} tool definitions")
+    
+    def register_tool(self, name: str, tool_class: Optional[type], hosted_by: str, description: str):
+        """Register a tool in the registry (metadata only)"""
+        self.tools[name] = {
+            "class": tool_class,  # Will be None for API-based tools
+            "hosted_by": hosted_by,
+            "description": description
+        }
+        logger.debug(f"Registered tool: {name} (hosted by {hosted_by})")
+    
+    def discover_tools(self, hosted_by: Optional[str] = None) -> List[Dict[str, Any]]:
+        """Discover available tools, optionally filtered by host"""
+        tools = []
+        for name, info in self.tools.items():
+            if hosted_by is None or info["hosted_by"] == hosted_by:
+                tools.append({
+                    "name": name,
+                    "hosted_by": info["hosted_by"],
+                    "description": info["description"]
+                })
+        return tools
+    
+    def execute_tool(self, tool_name: str, hosted_by: str, input_data: Dict[str, Any], 
+                    config: Optional[Dict[str, Any]] = None) -> Dict[str, Any]:
+        """Execute a tool - this should not be called directly in API-based mode"""
+        logger.warning(f"Direct tool execution attempted for {tool_name}. Use API client instead.")
+        return {
+            "success": False,
+            "error": "Direct tool execution not supported. Use API client for tool execution."
+        } 

memra-0.2.1/memra.egg-info/SOURCES.txt

@@ -0,0 +1,13 @@
+CHANGELOG.md
+LICENSE
+MANIFEST.in
+README.md
+pyproject.toml
+setup.py
+memra/__init__.py
+memra/discovery.py
+memra/discovery_client.py
+memra/execution.py
+memra/models.py
+memra/tool_registry.py
+memra/tool_registry_client.py

{memra-0.2.0 → memra-0.2.1}/pyproject.toml

@@ -4,12 +4,12 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "memra"
-version = "0.2.0"
-description = "Declarative framework for enterprise workflows with MCP integration"
+version = "0.2.1"
+description = "Declarative framework for enterprise workflows with MCP integration - Client SDK"
 readme = "README.md"
 license = {text = "MIT"}
 authors = [
-    {name = "Memra", email = "info@memra.co"}
+    {name = "Memra", email = "support@memra.com"}
 ]
 classifiers = [
     "Development Status :: 3 - Alpha",
@@ -29,7 +29,6 @@ dependencies = [
     "typing-extensions>=4.0.0",
     "aiohttp>=3.8.0",
     "aiohttp-cors>=0.7.0",
-    "psycopg2-binary>=2.9.0",
 ]
 
 [project.optional-dependencies]
@@ -39,6 +38,9 @@ dev = [
     "black",
     "flake8",
 ]
+mcp = [
+    "psycopg2-binary>=2.9.0",
+]
 
 [project.urls]
 Homepage = "https://memra.co"
@@ -48,4 +50,5 @@ Repository = "https://github.com/memra-platform/memra-sdk"
 memra = "memra.cli:main"
 
 [tool.setuptools.packages.find]
-include = ["memra*"] 
+include = ["memra*"]
+exclude = ["logic*", "server*", "app*", "mcp_bridge*"] 

{memra-0.2.0 → memra-0.2.1}/setup.py

@@ -5,10 +5,10 @@ with open("README.md", "r", encoding="utf-8") as fh:
 
 setup(
     name="memra",
-    version="0.2.0",
+    version="0.2.1",
     author="Memra",
-    author_email="info@memra.co",
-    description="Declarative framework for enterprise workflows with MCP integration",
+    author_email="support@memra.com",
+    description="Declarative framework for enterprise workflows with MCP integration - Client SDK",
     long_description=long_description,
     long_description_content_type="text/markdown",
     url="https://github.com/memra/memra-sdk",
@@ -31,7 +31,6 @@ setup(
         "typing-extensions>=4.0.0",
         "aiohttp>=3.8.0",
         "aiohttp-cors>=0.7.0",
-        "psycopg2-binary>=2.9.0",
     ],
     extras_require={
         "dev": [
@@ -40,6 +39,9 @@ setup(
             "black",
             "flake8",
         ],
+        "mcp": [
+            "psycopg2-binary>=2.9.0",
+        ],
     },
     entry_points={
         "console_scripts": [

memra-0.2.0/PKG-INFO

@@ -1,161 +0,0 @@
-Metadata-Version: 2.4
-Name: memra
-Version: 0.2.0
-Summary: Declarative framework for enterprise workflows with MCP integration
-Home-page: https://github.com/memra/memra-sdk
-Author: Memra
-Author-email: Memra <info@memra.co>
-License: MIT
-Project-URL: Homepage, https://memra.co
-Project-URL: Repository, https://github.com/memra-platform/memra-sdk
-Classifier: Development Status :: 3 - Alpha
-Classifier: Intended Audience :: Developers
-Classifier: License :: OSI Approved :: MIT License
-Classifier: Operating System :: OS Independent
-Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.8
-Classifier: Programming Language :: Python :: 3.9
-Classifier: Programming Language :: Python :: 3.10
-Classifier: Programming Language :: Python :: 3.11
-Requires-Python: >=3.8
-Description-Content-Type: text/markdown
-License-File: LICENSE
-Requires-Dist: pydantic>=1.8.0
-Requires-Dist: httpx>=0.24.0
-Requires-Dist: typing-extensions>=4.0.0
-Requires-Dist: aiohttp>=3.8.0
-Requires-Dist: aiohttp-cors>=0.7.0
-Requires-Dist: psycopg2-binary>=2.9.0
-Provides-Extra: dev
-Requires-Dist: pytest>=6.0; extra == "dev"
-Requires-Dist: pytest-asyncio; extra == "dev"
-Requires-Dist: black; extra == "dev"
-Requires-Dist: flake8; extra == "dev"
-Dynamic: author
-Dynamic: home-page
-Dynamic: license-file
-Dynamic: requires-python
-
-# Memra SDK
-
-A declarative orchestration framework for AI-powered business workflows. Think of it as "Kubernetes for business logic" where agents are the pods and departments are the deployments.
-
-## 🚀 Team Setup
-
-**New team member?** See the complete setup guide: **[TEAM_SETUP.md](TEAM_SETUP.md)**
-
-This includes:
-- Database setup (PostgreSQL + Docker)
-- Local development environment
-- Testing instructions
-- Troubleshooting guide
-
-## Quick Start
-
-```python
-from memra.sdk.models import Agent, Department, Tool
-
-# Define your agents
-data_extractor = Agent(
-    role="Data Extraction Specialist",
-    job="Extract and validate data",
-    tools=[Tool(name="DataExtractor", hosted_by="memra")],
-    input_keys=["input_data"],
-    output_key="extracted_data"
-)
-
-# Create a department
-dept = Department(
-    name="Data Processing",
-    mission="Process and validate data",
-    agents=[data_extractor]
-)
-
-# Run the workflow
-result = dept.run({"input_data": {...}})
-```
-
-## Installation
-
-```bash
-pip install memra
-```
-
-## API Access
-
-Memra requires an API key to execute workflows on the hosted infrastructure.
-
-### Get Your API Key
-Contact [info@memra.co](mailto:info@memra.co) for API access.
-
-### Set Your API Key
-```bash
-# Set environment variable
-export MEMRA_API_KEY="your-api-key-here"
-
-# Or add to your shell profile for persistence
-echo 'export MEMRA_API_KEY="your-api-key-here"' >> ~/.zshrc
-```
-
-### Test Your Setup
-```bash
-python examples/accounts_payable_client.py
-```
-
-## Documentation
-
-Documentation is coming soon. For now, see the examples below and in the `examples/` directory.
-
-## Example: Propane Delivery Workflow
-
-See the `examples/propane_delivery.py` file for a complete example of how to use Memra to orchestrate a propane delivery workflow.
-
-## Contributing
-
-We welcome contributions! Please see our [contributing guide](CONTRIBUTING.md) for details.
-
-## License
-
-This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
-
-## Examples
-
-```
-├── examples/
-│   ├── accounts_payable_client.py  # API-based example
-│   ├── accounts_payable.py         # Local example
-│   ├── invoice_processing.py       # Simple workflow
-│   └── propane_delivery.py         # Domain example
-├── memra/                  # Core SDK
-├── logic/                  # Tool implementations  
-├── local/dependencies/     # Database setup & schemas
-└── docker-compose.yml      # Database setup
-```
-
-## ✨ New: MCP Integration
-
-Memra now supports **Model Context Protocol (MCP)** integration, allowing you to execute operations on your local infrastructure while leveraging Memra's cloud-based AI processing.
-
-**Key Benefits:**
-- 🔒 **Keep sensitive data local** - Your databases stay on your infrastructure
-- ⚡ **Hybrid processing** - AI processing in the cloud, data operations locally  
-- 🔐 **Secure communication** - HMAC-authenticated requests between cloud and local
-- 🛠️ **Easy setup** - Simple bridge server connects your local resources
-
-**Quick Example:**
-```python
-# Agent that uses local database via MCP
-agent = Agent(
-    role="Data Writer",
-    tools=[{
-        "name": "PostgresInsert",
-        "hosted_by": "mcp",  # Routes to your local infrastructure
-        "config": {
-            "bridge_url": "http://localhost:8081",
-            "bridge_secret": "your-secret"
-        }
-    }]
-)
-```
-
-📖 **[Complete MCP Integration Guide →](docs/mcp_integration.md)**

memra-0.2.0/memra/__init__.py

@@ -1,24 +0,0 @@
-"""
-Memra SDK - Declarative framework for enterprise workflows with MCP integration
-
-A powerful orchestration framework that allows you to build AI-powered business workflows
-with hybrid cloud/local execution capabilities.
-"""
-
-__version__ = "0.2.0"
-__author__ = "Memra"
-__email__ = "info@memra.co"
-
-# Core imports
-from .models import Agent, Department, LLM, Tool
-from .execution import ExecutionEngine
-
-# Make key classes available at package level
-__all__ = [
-    "Agent",
-    "Department", 
-    "LLM",
-    "Tool",
-    "ExecutionEngine",
-    "__version__"
-] 

memra-0.2.0/memra/discovery_client.py

@@ -1,49 +0,0 @@
-"""
-Client-side tool discovery for Memra SDK
-Queries the Memra API to discover available tools
-"""
-
-from typing import List, Dict, Any, Optional
-from .tool_registry_client import ToolRegistryClient
-
-def discover_tools(hosted_by: Optional[str] = None) -> List[Dict[str, Any]]:
-    """
-    Discover available tools from the Memra API
-    
-    Args:
-        hosted_by: Filter tools by hosting provider ("memra" or "mcp")
-        
-    Returns:
-        List of available tools with their descriptions
-    """
-    registry = ToolRegistryClient()
-    return registry.discover_tools(hosted_by)
-
-def check_api_health() -> bool:
-    """
-    Check if the Memra API is available
-    
-    Returns:
-        True if API is healthy, False otherwise
-    """
-    registry = ToolRegistryClient()
-    return registry.health_check()
-
-def get_api_status() -> Dict[str, Any]:
-    """
-    Get detailed API status information
-    
-    Returns:
-        Dictionary with API status details
-    """
-    registry = ToolRegistryClient()
-    
-    is_healthy = registry.health_check()
-    tools = registry.discover_tools() if is_healthy else []
-    
-    return {
-        "api_healthy": is_healthy,
-        "api_url": registry.api_base,
-        "tools_available": len(tools),
-        "tools": tools
-    }