Merge remote-tracking branch 'origin/main' into efs-cleanup-testing

Author: remi-td

README.md

@@ -37,6 +37,13 @@ We are providing groupings of tools and associated helpful prompts to support al
 
 **Step 3.** - There are many client options availale, the [Client Guide](docs/client_guide/CLIENT_GUIDE.md) explains how to configure and run a sample of different clients.
 
+<br>
+
+[A Video Library](./docs/VIDEO_LIBRARY.md) has been curated to assist.
+
+<br>
+
+
 
 ### Quick start with Claude desktop
 If you want to quickly evaluate the tool, we recommend using Claude desktop, the uv package manager and [Teradata Clearscape Experience](https://www.teradata.com/getting-started/demos/clearscape-analytics).
@@ -65,6 +72,10 @@ If you want to quickly evaluate the tool, we recommend using Claude desktop, the
   }
 ```
 
+## Contributing
+Please refer to the [Contributing](./docs/CONTRIBUTING.md) guide and the [Developer Guide](./docs/developer_guide/DEVELOPER_GUIDE.md).
+
+
 ---------------------------------------------------------------------
 ## Certification
 <a href="https://glama.ai/mcp/servers/@Teradata/teradata-mcp-server">

docs/CONTRIBUTING.md

@@ -12,4 +12,4 @@ Make sure you have setup your environment based on the Developer Guide in this r
 - Code structure should be maintained. The structure of the repo and files has been carefully crafted, and any deviations from that should be only done when agreed upon by the entire team.
 
 ## Design Guidelines
-- Refer to Developer Guide.
+- Refer to [Developer Guide](./developer_guide/DEVELOPER_GUIDE.md).

docs/GETTING_STARTED.md

@@ -1,6 +1,6 @@
 # MCP Server - Getting Started
 
-![Getting Started](media/MCP.png)
+![Getting Started](media/MCP-quickstart.png)
 
 This document will cover the process and options for getting the teradata-mcp-server up and running
 
@@ -10,8 +10,10 @@ Assumes that you have a running Teradata environment, you should have the follow
 3. user password - password for the corresponding user name
 4. database - On Teradata systems this is typically the same as you user name
 
-## Step 0 - Installing git
-Refer to the [git](https://git-scm.com/) website for download and installation instructions for your environment.
+## Step 0 - Installing environment
+- Refer to the [git](https://git-scm.com/) website for download and installation instructions, git will be used to pull code from the github repository.
+- Refer to the [Installing uv](https://docs.astral.sh/uv/getting-started/installation/#standalone-installer) guide to assist with installing uv, uv is used to manage packages that are used.
+- Refer to the [Python.org](https://www.python.org/downloads/) for downloading and installing python 3.11 or greater.
 
 ## Step 1 - Download the Software
 Clone the mcp-server repository with: 
@@ -51,14 +53,9 @@ Deployment Choice
 
 The recommended choice will be to deploy Streamable-http in a docker container. [Jump to next section](#step-5---using-docker) for the docker option.
 
-## Step 4 - Decide on the tools/prompts you want to be available (optional) to a profile
-
-Open the [configure_tools.yaml](../configure_tools.yml) file.  The first level of the structure is the profile name, then the module name, then allmodule/tools/prompts
-
-For each module if you set allmodule to True then all tools and prompts will be visible unless you explicitly set the tool/prompt name to False.  If the allmodule is set to False, then nothing will be visible unless you explicitly set the tool/prompt name to True.
-
-If no profile is defined then the default profile will be "all" this will enable all tools and prompts.
+## Step 4 - Customizing the server (optional)
 
+Refer to the [Customizing](CUSTOMIZING.md) instuctions 
 
 --------------------------------------------------------------------------------------
 ## Step 5 - Using Docker

docs/VIDEO_LIBRARY.md

@@ -0,0 +1,12 @@
+# Video Library
+
+The video library is a curation of videos that the community believes would be valuable for new members.
+
+## Demo's
+[DBA use cases](https://www.youtube.com/watch?v=CYUs1tRoO1M&t=14s) - this demo is a short overview of the server along with some simple DBA use cases.
+
+
+## Installation
+[Installing on a mac](https://youtu.be/bx6q3iL6K4A) - this is a longer video walking you through the steps to install the server on your mac.
+
+

docs/client_guide/CLIENT_GUIDE.md

@@ -1,6 +1,6 @@
 # Working with MCP Clients
 
-![Getting Started](../media/MCP.png)
+![Getting Started](../media/MCP-quickstart.png)
 
 This documet will cover the process and options for getting a client tool to connect to the teradata-mcp-server.  Note that you have many client options, we will cover some but not all.
 
@@ -10,19 +10,26 @@ The transport mode that the server is running needs to be the same as the client
 
 If you configured the server as a Streamable-http transport mode then the client also needs to be configured as a streamable-http mode.
 
+There are many client options that you can choose from, they each have different strengths, we have curated a small number so that you have examples.
 
-## Clients
+## GUI Clients
 
 - [Claude Desktop](./Claude_desktop.md)
 - [Microsoft Copilot](./Microsoft_copilot.md)
 - [Google AgentSpace](./Google_agentspace.md)
+- [Open WebUI](./Open_WebUI.md)
+- [teradata-trusted-data-agent](https://github.com/rgeissen/teradata-trusted-data-agent)
+
+## Audio Client
+- [MCP Voice Client](../../test/MCP_VoiceClient/README.md)
+
+## Code Clients
 - [Visual Studio Code](./Visual_Studio_Code.md)
 - [MCP Inspector](./MCP_Inspector.md) - For testing MCP server
 - [Code Agents](./Code_Agents.md) - Google ADK Agent & Pydanicai Agent
 - [REST API](./Rest_API.md)
-- [Open WebUI](./Open_WebUI.md)
 - [Google Gemini CLI](./Google_Gemini_CLI.md)
-- [MCP Web Client](../../test/MCP_WebClient/README.md)
+
 
 
 

docs/developer_guide/DEVELOPER_GUIDE.md

@@ -123,7 +123,14 @@ Two guides have been created to show how to add tools and prompts:
 <br>
 
 ## Tools testing
-Every module will have at least one prompt that would be responsible for testing all of the tools in the module.  The following template should be used in the module_objects.ylm file, the name of the testig prompt should be test_\<module name\>Tools:
+Every module will have at least one prompt that would be responsible for testing all of the tools in the module.  
+
+Use the tester profile to see the test prompts.
+
+`uv run teradata-mcp-server --profile tester`
+
+The following test prompt template should be modified and placed in the module_objects.ylm file (e.g. base_objects.yml), the name of the testig prompt should be test_\<module name\>Tools. (e.g. test_baseTools):
+
 
 ```
 test_baseTools:
@@ -136,39 +143,54 @@ test_baseTools:
    
    ## Phase 0 - Get a list of databases
    - Get a list of databases from the user. Use the base_databaseList function to get the list.  The tool should return a list of databases in the Teradata system.
+
    ## Phase 1 - Get the list of tables in each database
    - For the DBC database, get the list of tables. Use the base_tableList function to get the list. The tool should return a list of tables in the DBC database.
+
    ## Phase 2 - Create a test table in your default database
    - Create a customer table in your default database called test_customer, the table should have an Cust_id column. Use the base_writeQuery function to create the table. A test_customer table should be created.
    - Add 10 rows to the test_customer table. Use the base_writeQuery function to add the rows.
+
    ## Phase 3 - Test the query tool
    - test the query tool by running a query on the test_customer table. Use the base_readQuery function to run the query. The query should return the 10 rows you added in the previous phase.
    - test the query tool by running a query on the test_customer table with a filter. Use the base_readQuery function to run the query. The query should return only the rows that match the filter. (example of a filter is Cust_id > 5)
+
    ## Phase 4 - Test the table DDL tool
    - Get the DDL of the test_customer table. Use the base_tableDDL function to get the DDL. The tool should return the DDL of the test_customer table.
+
    ## Phase 5 - Test the column description tool
    - Get the column description of the test_customer table. Use the base_columnDescription function to get the columns description. The tool should return the column description of the test_customer table.
+
    ## Phase 6 - Test the table preview tool
    - Get the preview of the test_customer table. Use the base_tablePreview function to get the preview. The tool should return the first 5 rows of the test_customer table and the column information.
+
    ## Phase 7 - Test the table affinity tool
    - Get the table affinity of the test_customer table. Use the base_tableAffinity function to get the affinity. The tool should return the table affinity of the test_customer table.
+
    ## Phase 8 - Test the table usage tool
    - Get the table usage of the test_customer table. Use the base_tableUsage function to get the usage. The tool should return the table usage of the test_customer table.
+
    ## Phase 9 - Clean up
    - Drop the test_customer table. Use the base_writeQuery function to drop the table. The tool should return a success message.
+
    ## Communication guidelines:
          - Be concise but informative in your explanations
          - Clearly indicate which phase the process is currently in
-         - summarize the success of the phase before moving to the next phase
+         - summarize the success or any failure of the phase before moving to the next phase
    ## Final output guidelines:
          - return in markdown results for all phases
          - Example:
          ***Phase 0:*** list databases:  Successful
          ***Phase 1:*** list tables in DBC database:  Successful
-         ***Phase 2:*** Create test_customer table:  Successful
+         ***Phase 2:*** Create test_customer table:  Failed
 
 ```
 
+Before code is submitted, all test scripts should run with a successful outcome, this ensures that new code does not break old code.  All new code needs to have test prompts added to the module test prompt.
+
+You can run the test prompts via your client of choice.
+
+
 <br><br><br>
 
 # Development Cycle

profiles.yml

@@ -1,4 +1,12 @@
 all:
+  tool:
+    - .*
+  prompt:
+    - ^(?!test_).*
+  resource:
+    - .*
+
+tester:
   tool:
     - .*
   prompt:
@@ -8,20 +16,20 @@ all:
 
 dba:
   tool:
-    - dba_*
-    - base_*
-    - sec_*
+    - ^dba_*
+    - ^base_*
+    - ^sec_*
   prompt:
-    - dba_*
+    - ^dba_*
 
 dataScientist:
   tool:
-    - base_*
-    - rag_*
-    - fs_*
-    - qlty_*
-    - sec_userDbPermissions
-    - dba_userSqlList
+    - ^base_*
+    - ^rag_*
+    - ^fs_*
+    - ^qlty_*
+    - ^sec_userDbPermissions
+    - ^dba_userSqlList
 
 eda:
   tool:

src/teradata_mcp_server/tools/base/base_objects.yml

@@ -22,37 +22,47 @@ test_baseTools:
    Perform the phases in order, and do not skip any phase.
    
    ## Phase 0 - Get a list of databases
-   - Get a list of databases from the user. Use the base_databaseList function to get the list.  The tool should return a list of databases in the Teradata system.
+   - Get a list of databases from the user. Use the base_databaseList function to get the list.  The tool should return a list of databases in the Teradata system, if no data is returned then fail this test.
    ## Phase 1 - Get the list of tables in each database
-   - For the DBC database, get the list of tables. Use the base_tableList function to get the list. The tool should return a list of tables in the DBC database.
+   - For the DBC database, get the list of tables. Use the base_tableList function to get the list. The tool should return a list of tables in the DBC database, if no data is returned then fail this test.
    ## Phase 2 - Create a test table in your default database
    - Create a customer table in your default database called test_customer, the table should have an Cust_id column. Use the base_writeQuery function to create the table. A test_customer table should be created.
    - Add 10 rows to the test_customer table. Use the base_writeQuery function to add the rows.
    ## Phase 3 - Test the query tool
-   - test the query tool by running a query on the test_customer table. Use the base_readQuery function to run the query. The query should return the 10 rows you added in the previous phase.
-   - test the query tool by running a query on the test_customer table with a filter. Use the base_readQuery function to run the query. The query should return only the rows that match the filter. (example of a filter is Cust_id > 5)
+   - test the query tool by running a query on the test_customer table. Use the base_readQuery function to run the query. The query should return the 10 rows you added in the previous phase, if no data is returned then fail this test.
+   - test the query tool by running a query on the test_customer table with a filter. Use the base_readQuery function to run the query. The query should return only the rows that match the filter, if no data is returned then fail this test. (example of a filter is Cust_id > 5)
    ## Phase 4 - Test the table DDL tool
-   - Get the DDL of the test_customer table. Use the base_tableDDL function to get the DDL. The tool should return the DDL of the test_customer table.
+   - Get the DDL of the test_customer table. Use the base_tableDDL function to get the DDL. The tool should return the DDL of the test_customer table, if no data is returned then fail this test.
    ## Phase 5 - Test the column description tool
-   - Get the column description of the test_customer table. Use the base_columnDescription function to get the columns description. The tool should return the column description of the test_customer table.
+   - Get the column description of the test_customer table. Use the base_columnDescription function to get the columns description. The tool should return the column description of the test_customer table, if no data is returned then fail this test.
    ## Phase 6 - Test the table preview tool
-   - Get the preview of the test_customer table. Use the base_tablePreview function to get the preview. The tool should return the first 5 rows of the test_customer table and the column information.
+   - Get the preview of the test_customer table. Use the base_tablePreview function to get the preview. The tool should return the first 5 rows of the test_customer table and the column information, if no data is returned then fail this test.
    ## Phase 7 - Test the table affinity tool
-   - Get the table affinity of the test_customer table. Use the base_tableAffinity function to get the affinity. The tool should return the table affinity of the test_customer table.
+   - Get the table affinity of the test_customer table. Use the base_tableAffinity function to get the affinity. The tool should return the table affinity of the test_customer table, if no data is returned then fail this test.
    ## Phase 8 - Test the table usage tool
-   - Get the table usage of the test_customer table. Use the base_tableUsage function to get the usage. The tool should return the table usage of the test_customer table.
+   - Get the table usage of the test_customer table. Use the base_tableUsage function to get the usage. The tool should return the table usage of the test_customer table, if no data is returned then fail this test.
    ## Phase 9 - Clean up
    - Drop the test_customer table. Use the base_writeQuery function to drop the table. The tool should return a success message.
    ## Communication guidelines:
          - Be concise but informative in your explanations
          - Clearly indicate which phase the process is currently in
-         - summarize the success of the phase before moving to the next phase
-   ## Final output guidelines:
-         - return in markdown results for all phases
-         - Example:
-         ***Phase 0:*** list databases:  Successful
-         ***Phase 1:*** list tables in DBC database:  Successful
-         ***Phase 2:*** Create test_customer table:  Successful
+         - summarize the success or any failure of the phase before moving to the next phase
+    ## Final output guidelines:
+          - return in markdown results for all phases as a table including the phase number, description, success or failure status, and a description of the outcome.
+          - if a tool does not return any results, then fail that test and indicate it in the outcome.
+    ## Example output:
+          - | Phase | Description | Status | Outcome |
+          - |-------|-------------|--------|---------|
+          - | 0     | Get a list of databases | Success | List of databases returned |
+          - | 1     | Get the list of tables in DBC database | Success | List of tables returned |
+          - | 2     | Create a test table in default database | Success | test_customer table created |
+          - | 3     | Test the query tool on test_customer table | Success | 10 rows returned |
+          - | 4     | Get the DDL of the test_customer table | Success | DDL returned |
+          - | 5     | Get the column description of the test_customer table | Success | Column description returned |
+          - | 6     | Get the preview of the test_customer table | Success | First 5 rows and column info returned |
+          - | 7     | Get the table affinity of the test_customer table | Success | Table affinity returned |
+          - | 8     | Get the table usage of the test_customer table | Success | Table usage returned |
+          - | 9     | Drop the test_customer table | Success | Table dropped successfully |
 
          
 base_query: