Lab setup

Your environment has already been pre-configured for you.

Your Showroom is split into two parts. You are reading these instructions on the left. On the right hand side you see four tabs. On each tab log into the application to prepare for the demo/lab.

Your user’s password is the same for all applications.

OpenShift Console

  1. On the right panel switch to the OpenShift Console tab if it’s not already selected.

  2. Log in using user {user} and password {password}

  3. Upon seeing the OpenShift Console click on Skip tour and then minimize the Lightspeed popup if it appears.

LibreChat

  1. On the right panel switch to the LibreChat tab.

  2. Log in using your LibreChat user: {librechat_user} with password {librechat_password}

  3. Activate the large language model, that you want to use, by clicking on the My Agents button top left. Select LiteMAAS → llama-scout-17b.

  4. There are two pre-deployed MCP Servers in your environment. One to interact with OpenShift and one to interact with Gitea.

    1. In your chat text window click on the MCP Servers button and select both the openshift and gitea MCP Servers.

    2. Note that there are two additional MCP Servers defined: fetch and yardstick - you will deploy those much later in the lab. Do not attempt to activate those at this time.

    3. For the two that we are activating, if you see an orange tool icon next to the name, make sure to click that and initialize the MCP Servers. If the servers are not initialized the LLM can’t use them.

  5. The MCP Servers dropdown label should change to read 2 selected. This means that your model can now use both predeployed MCP Servers.

    If you had to initialize the MCP Servers that means that the agent pod was not able to initialize the server connection either. This can happen if the environment had been stopped between the time of deployment and when you are using this environment.

    To fix this, switch to the OpenShift Console, find the agent-{user} namespace and delete the agent pod (its name starts with agent-). This will re-initialize the LLM and MCP Server connection.

    You can double check the logs of the recreated pod where you should see that both MCP Servers have been connected successfully:

    INFO: Started server process [1]
INFO: Waiting for application startup.
============================================================
🚀 Pipeline Failure Agent - Starting up
============================================================
📋 Loading MCP configuration from environment...
OpenShift: http://mcp-openshift-proxy.mcp-openshift-user4.svc.cluster.local:8080/sse#openshift (sse)
Gitea: http://mcp-gitea-proxy.mcp-gitea-user4.svc.cluster.local:8080/mcp (streamable-http)
🔌 Initializing MCP connections (2 servers)...
Connecting to openshift (sse)...
URL: http://mcp-openshift-proxy.mcp-openshift-user4.svc.cluster.local:8080/sse#openshift
SSE message endpoint: http://mcp-openshift-proxy.mcp-openshift-user4.svc.cluster.local:8080/sse?sessionid=HSUO2GH5V5W22VT4Z3BDARLALT
SSE init result: {'jsonrpc': '2.0', 'id': 1, 'result': {'capabilities': {'logging': {}, 'prompts': {'listChanged': True}, 'tools': {'listChanged': True}}, 'protocolVersion': '2024-11-05', 'serverInfo': {'name': 'kubernetes-mcp-server', 'title': 'kubernetes-mcp-server', 'version': 'v0.0.56'}}}
✅ openshift: connected
Tools (23): ['configuration_view', 'events_list', 'helm_install', 'helm_list', 'helm_uninstall', 'namespaces_list', 'nodes_log', 'nodes_stats_summary', 'nodes_top', 'pods_delete', 'pods_exec', 'pods_get', 'pods_list', 'pods_list_in_namespace', 'pods_log', 'pods_run', 'pods_top', 'projects_list', 'resources_create_or_update', 'resources_delete', 'resources_get', 'resources_list', 'resources_scale']
Connecting to gitea (streamable-http)...
URL: http://mcp-gitea-proxy.mcp-gitea-user4.svc.cluster.local:8080/mcp
✅ gitea: connected
Tools (106): ['add_issue_labels', 'add_tracked_time', 'cancel_repo_action_run', 'clear_issue_labels', 'create_branch', 'create_file', 'create_issue', 'create_issue_comment', 'create_milestone', 'create_org_action_variable', 'create_org_label', 'create_pull_request', 'create_pull_request_review', 'create_pull_request_reviewer', 'create_release', 'create_repo', 'create_repo_action_variable', 'create_repo_label', 'create_tag', 'create_wiki_page', 'delete_branch', 'delete_file', 'delete_milestone', 'delete_org_action_secret', 'delete_org_action_variable', 'delete_org_label', 'delete_pull_request_review', 'delete_pull_request_reviewer', 'delete_release', 'delete_repo_action_secret', 'delete_repo_action_variable', 'delete_repo_label', 'delete_stopwatch', 'delete_tag', 'delete_tracked_time', 'delete_wiki_page', 'dismiss_pull_request_review', 'dispatch_repo_action_workflow', 'download_repo_action_job_log', 'edit_issue', 'edit_issue_comment', 'edit_milestone', 'edit_org_label', 'edit_repo_label', 'fork_repo',...
✅ Startup complete - 129 tools available
============================================================
INFO: Application startup complete.
INFO: Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)

    This is a limitation of the rather simple Python code that makes up our agent - a real agent would have logic to detect that it can’t connect to the MCP Servers and then periodically retry. But for the purposes of this lab the code is good enough.

Gitea

This lab environment uses Gitea to store all GitOps and source code artifacts. The reason we are using Gitea and not GitHub or GitLab is that it is much simpler to set up - and we are able to pre-configure the users and repositories automatically - no forking repositories, creating tokens, etc. required.

  1. On the right panel switch to the Gitea tab.

  2. Click on Sign in in the top right corner and log in using your Gitea user: {gitea_user} with password {gitea_password}

OpenShift GitOps

Your applications are deployed using OpenShift Gitops from your Gitea repository. If you need to check the sync status later in the lab you can use this tab to log into the OpenShift GitOps console. This should not be necessary if you follow the instructions exactly.

  1. On the right panel switch to the OpenShift GitOps tab if it’s not already selected.

  2. Click on Log In Via OpenShift.

  3. Log in using user {user} and password {password}.

  4. On the Authorize Access screen click Allow selected permissions.

Switch back to the OpenShift Console tab to be ready for the demo / lab scenario.