- name
- ums-user-management
- description
- TODO: Write a concise description explaining the skill's purpose — managing users in the UMS via CRUD operations, search, web enrichment; and when to activate it
- license
- Apache-2.0
- metadata
- {"author":"ai-powered-apps-development-expert","version":"2.0"}
# UMS User Management
<!-- TODO: Describe the agent's role and name the two MCP servers it has access to -->
---
## MCP Server Connections
<!-- TODO:
Add a Markdown table with three columns: Server | Transport | URL
- UMS MCP Server → streamable-http, http://localhost:8005/mcp
- DuckDuckGo Search → streamable-http, http://localhost:8000/mcp -->
---
## Available MCP Tools
### UMS MCP Server Tools
<!-- TODO:
Add a table listing all five UMS tools with their description and key parameters:
get_user_by_id, search_user, add_user, update_user, delete_user
Below the table document:
- UserCreate required fields (name, surname, email, about_me)
- UserCreate optional fields (phone, date_of_birth, address, gender, company, salary, credit_card)
- UserSearchRequest fields (all optional, partial case-insensitive match except gender)
- UserUpdate fields (same optional set as UserCreate) -->
---
### DuckDuckGo Search MCP Server Tools
<!-- TODO:
Add a table listing both DuckDuckGo tools with parameters:
- search: query, max_results (default 10, max 50)
- fetch_content: url (must start with http:// or https://)
Describe when to use search vs fetch_content -->
---
## Operating Rules
<!-- TODO:
Number the operating rules:
1. Always explain actions before executing any tool call
2. UMS-first: always query UMS before resorting to web search
3. Web search for enrichment when user data is incomplete or ambiguous
4. Confirm before creating: present the full proposed profile and wait for confirmation
5. Deletions require confirmation: warn that the action is permanent and irreversible
6. Format responses clearly and structurally
7. Handle errors gracefully: explain what went wrong and suggest alternatives -->
---
## Workflows
### Finding a User
<!-- TODO:
Step-by-step workflow:
1. Call search_user with available criteria
2. If results found → present them to the operator
3. If no results → inform the operator; offer web search if context suggests a real person -->
### Adding a User
<!-- TODO:
Step-by-step workflow:
1. Collect available user data from the operator
2. Identify missing required fields (name, surname, email, about_me)
3. If data incomplete:
a. Call search (DuckDuckGo) with the person's name/company/context
b. Optionally call fetch_content on a relevant URL for deeper detail
c. Build a complete UserCreate profile from gathered data
d. Present the full profile to the operator for confirmation
4. On confirmation → call add_user -->
### Updating a User
<!-- TODO:
Step-by-step workflow:
1. If user_id unknown → call search_user to locate the user first
2. Confirm which fields to update with the operator
3. Call update_user with only the fields that need to change
4. Report success or explain any error -->
### Deleting a User
<!-- TODO:
Step-by-step workflow:
1. If user_id unknown → call search_user to locate the user first
2. Display user details and warn: deletion is permanent and cannot be undone
3. Wait for explicit operator confirmation
4. On confirmation → call delete_user
5. Report success or explain any error -->
---
## Boundaries
<!-- TODO: Define the agent's scope — user management only; politely redirect off-topic requests -->
Auf GitHub ansehen