WORKSPACE MULTI ENDPOINTS (Version 0.236.045)

Overview

Workspace multi-endpoint management extends the admin multi-endpoint system to personal and group workspaces. Users can configure workspace endpoints that live alongside global endpoints, and agent model selection is driven by these combined lists.

Implemented in version: 0.236.045

Updated in version: 0.242.071

Dependencies

• Global model endpoints configured in admin settings
• Workspace endpoint storage in user settings and group documents
• Agent modal updates for multi-endpoint and Foundry agent discovery

Technical Specifications

Architecture Overview

• Global endpoints remain in application settings.
• Personal endpoints are stored in user settings under personal_model_endpoints.
• Group endpoints are stored on group documents under model_endpoints.
• Agent modal requests a combined, sanitized endpoint list for model selection.
• Foundry agent lookup uses endpoint IDs to resolve authentication and list agents.

API Endpoints

• GET /api/user/model-endpoints / POST /api/user/model-endpoints
• GET /api/group/model-endpoints / POST /api/group/model-endpoints
• GET /api/user/agent/settings
• GET /api/group/agent/settings
• POST /api/models/foundry/agents
• POST /api/user/models/fetch / POST /api/user/models/test-model
• POST /api/group/models/fetch / POST /api/group/models/test-model

Configuration

• Global toggle: enable_multi_model_endpoints in application/single_app/config.py
• Workspace endpoints stored per user and per group

File Structure

• Frontend templates: application/single_app/templates/workspace.html, application/single_app/templates/group_workspaces.html, application/single_app/templates/_agent_modal.html
• Frontend logic: application/single_app/static/js/workspace/workspace_model_endpoints.js, application/single_app/static/js/agent_modal_stepper.js
• Backend: application/single_app/route_backend_models.py, application/single_app/route_backend_agents.py, application/single_app/semantic_kernel_loader.py

Usage Instructions

Enable/Configure

1. Admin enables multi-endpoint model management in admin settings.
2. Users open Personal Workspace or Group Workspace and add endpoints under the new Workspace/Group Model Endpoints card.
3. In the agent modal, select a model from the combined endpoint list.

Use the Setup Guide button in the endpoint table or Model Endpoint modal for in-product RBAC reminders. For Azure OpenAI, Foundry (classic), or New Foundry managed identity and service principal setup, see Configure Model Endpoint Identity. The same RBAC guidance applies to global, personal, and group-scoped endpoints.

Foundry Agent Import

1. Select an Azure AI Foundry endpoint in the Foundry section of the agent modal.
2. Click Fetch Agents to list available agents.
3. Select an agent to import its identity, then save.

User Workflows

• Personal agents can select models from global + personal endpoints.
• Group agents can select models from global + group endpoints.
• Foundry agents auto-populate identities from the selected Foundry endpoint.

Testing and Validation

• Functional test: functional_tests/test_workspace_multi_endpoints.py
• Manual validation:
  • Add workspace endpoints and ensure they appear in agent model dropdowns.
  • Verify Foundry agent list import using configured endpoints.

Performance Considerations

• Model discovery uses on-demand API calls to Azure/Foundry endpoints.

Known Limitations

• Workspace endpoints require configured credentials; only stored secrets are used for runtime resolution.