[FEATURE] Implement MCP prompts for guided workflows #7
Description
Implement MCP Prompts
Feature Description
Implement MCP prompts to provide guided, interactive workflows for common Odoo operations. Prompts help users through complex tasks by asking for information step-by-step, validating inputs, and suggesting options.
Motivation
Many Odoo operations require multiple fields and have complex validation rules. Prompts can:
- Guide users through required fields
- Validate data before submission
- Suggest related records
- Provide context-aware help
- Reduce errors and improve data quality
Proposed Prompts
1. Customer Creation Prompt
@mcp.prompt()
async def create_customer_prompt() -> Dict[str, Any]:
"""Interactive prompt for creating a new customer."""
return {
"name": "create_customer",
"description": "Create a new customer with guided input",
"parameters": [
{
"name": "company_name",
"description": "Company name (leave empty for individual)",
"type": "string",
"required": False
},
{
"name": "contact_name",
"description": "Contact person name",
"type": "string",
"required": True
},
{
"name": "email",
"description": "Email address",
"type": "string",
"pattern": r"^[\w\.-]+@[\w\.-]+\.\w+$",
"required": True
},
{
"name": "phone",
"description": "Phone number",
"type": "string",
"required": False
},
{
"name": "country",
"description": "Country",
"type": "selection",
"options": await self._get_country_options(),
"required": True
},
{
"name": "tags",
"description": "Customer tags",
"type": "multi-selection",
"options": await self._get_tag_options(),
"required": False
}
],
"on_submit": "create_customer_from_prompt"
}2. Product Search Assistant
@mcp.prompt()
async def product_search_prompt() -> Dict[str, Any]:
"""Smart product search with filters."""
return {
"name": "product_search",
"description": "Search products with smart filters",
"parameters": [
{
"name": "search_term",
"description": "Product name or reference",
"type": "string",
"required": False
},
{
"name": "category",
"description": "Product category",
"type": "selection",
"options": await self._get_category_tree(),
"required": False
},
{
"name": "price_range",
"description": "Price range",
"type": "range",
"min": 0,
"max": 10000,
"step": 10,
"required": False
},
{
"name": "in_stock_only",
"description": "Only show products in stock",
"type": "boolean",
"default": True
},
{
"name": "attributes",
"description": "Product attributes",
"type": "dynamic",
"handler": "get_attribute_filters"
}
],
"preview": True, # Show results preview before final selection
"on_submit": "search_products_from_prompt"
}3. Data Import Wizard
@mcp.prompt()
async def import_data_prompt() -> Dict[str, Any]:
"""Import data from CSV/Excel with field mapping."""
return {
"name": "import_data",
"description": "Import records from file with field mapping",
"steps": [
{
"name": "file_selection",
"parameters": [{
"name": "file_content",
"description": "Paste CSV content or file path",
"type": "text",
"required": True
}]
},
{
"name": "model_selection",
"parameters": [{
"name": "model",
"description": "Target Odoo model",
"type": "selection",
"options": await self._get_writable_models(),
"required": True
}]
},
{
"name": "field_mapping",
"parameters": "dynamic", # Generated based on CSV headers
"handler": "generate_field_mapping"
},
{
"name": "validation_options",
"parameters": [
{
"name": "skip_errors",
"description": "Skip rows with errors",
"type": "boolean",
"default": True
},
{
"name": "update_existing",
"description": "Update existing records",
"type": "boolean",
"default": False
}
]
}
],
"on_submit": "import_data_from_prompt",
"supports_progress": True
}4. Relationship Explorer
@mcp.prompt()
async def explore_relationships_prompt() -> Dict[str, Any]:
"""Explore record relationships interactively."""
return {
"name": "explore_relationships",
"description": "Navigate through record relationships",
"parameters": [
{
"name": "starting_point",
"description": "Start from model/record",
"type": "model_record_selector",
"required": True
},
{
"name": "relationship_depth",
"description": "How many levels to explore",
"type": "integer",
"min": 1,
"max": 5,
"default": 2
},
{
"name": "relationship_types",
"description": "Types of relationships to follow",
"type": "multi-selection",
"options": [
"many2one",
"one2many",
"many2many"
],
"default": ["many2one", "one2many"]
}
],
"interactive": True, # Allow drilling down during exploration
"on_submit": "explore_from_prompt"
}Implementation Requirements
Prompt Registration
# In prompts.py
class OdooMCPPrompts:
def __init__(self, server: 'OdooMCPServer'):
self.server = server
self.app = server.app
self._register_prompts()
def _register_prompts(self):
"""Register all prompts with FastMCP."""
self.app.prompt(self.create_customer_prompt)
self.app.prompt(self.product_search_prompt)
self.app.prompt(self.import_data_prompt)
self.app.prompt(self.explore_relationships_prompt)Validation Framework
class PromptValidator:
"""Validate prompt inputs before processing."""
async def validate_email(self, value: str) -> bool:
"""Validate email format."""
pattern = r"^[\w\.-]+@[\w\.-]+\.\w+$"
return bool(re.match(pattern, value))
async def validate_phone(self, value: str) -> bool:
"""Validate phone number format."""
# Implementation depends on country
pass
async def validate_selection(self, value: str, options: List[str]) -> bool:
"""Ensure selection is from allowed options."""
return value in optionsProgress Support
class ProgressReporter:
"""Report progress for long operations."""
async def report_import_progress(self, current: int, total: int, message: str):
"""Send progress update to client."""
await self.app.report_progress({
"operation": "import_data",
"current": current,
"total": total,
"percentage": (current / total) * 100,
"message": message
})Success Criteria
- At least 4 prompts implemented
- Input validation works correctly
- Dynamic options loaded from Odoo
- Progress reporting for long operations
- Error handling with helpful messages
- Integration tests for each prompt
- Documentation with examples
- Prompt discovery through MCP protocol
Benefits
- User Experience: Guided workflows reduce errors
- Data Quality: Validation ensures clean data
- Efficiency: Batch operations with progress tracking
- Discoverability: Users can explore available operations
- Integration: Works with any MCP-compatible client
Testing Requirements
- Unit tests for validation logic
- Integration tests for each prompt
- Test dynamic option loading
- Test progress reporting
- Test error scenarios
- Test with various MCP clients
Documentation
Create documentation covering:
- How to use each prompt
- Examples of prompt interactions
- How to extend with custom prompts
- Integration with MCP clients
- Best practices for prompt design
Priority
Medium - While not critical for basic functionality, prompts significantly improve user experience and reduce errors in complex operations.