fix: dev-v4 to main branch downmerge by Dhruvkumar-Microsoft · Pull Request #1098 · microsoft/Multi-Agent-Custom-Automation-Engine-Solution-Accelerator

Dhruvkumar-Microsoft · 2026-07-17T04:31:59Z

Purpose

This pull request updates the platform to use the new GPT-5.4 model as the default language model across deployment, configuration, and documentation. It also updates related workflow environment variables, quota checks, and deployment settings to align with the new model naming. Additionally, it provides improved post-deployment instructions in the developer experience scripts.

Model upgrade and configuration changes:

All references to gpt-4.1-mini are replaced with gpt-5.4-mini as the default deployment name in agent team configuration files (e.g., contract_compliance_team.json, hr.json), and "is_default": true is set for the new model. [1] [2] [3] [4] [5]
Workflow environment variables and deployment parameters are renamed from GPT_MIN_CAPACITY, O4_MINI_MIN_CAPACITY, and GPT41_MINI_MIN_CAPACITY to GPT_5_4_MINI_MIN_CAPACITY and GPT_5_4_MIN_CAPACITY throughout all relevant GitHub Actions workflows, with corresponding capacity values updated. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11]

Documentation and developer experience:

The default LLM in the transparency FAQ is updated from GPT-4o to GPT-5.4, with references and recommendations adjusted accordingly.
Post-deployment instructions in azure.yaml are clarified and expanded for both PowerShell and Bash, guiding users to build and push container images. [1] [2] [3]

Workflow cleanup:

The .github/workflows/create-release.yml workflow is removed, indicating a possible change in release automation strategy.

Does this introduce a breaking change?

Yes
No

How to Test

Get the code

git clone [repo-address]
cd [repo-name]
git checkout [branch-name]
npm install

Test the code

What to Check

Verify that the following are valid

...

Other Information

fix: Dev v4 to main PR

chore: delete Create-Release workflow due to codfish security incident

@Vinay-Microsoft

Replace @Vinay-Microsoft -> @VinaySh-Microsoft and @Prajwal-Microsoft -> @Prajwal1-Microsoft

fix: remove sensitive cookie data from sample user configuration

…ipts

…e data after image build and push

… JSON templates. - Updated deployment capacity defaults for GPT models. - Added private endpoint configurations for Azure Container Registry based on networking settings. - Adjusted scripts for building and pushing images to temporarily relax ACR restrictions during WAF deployments. - Enhanced WebSocket handling in the frontend to coalesce streaming messages for improved UI performance. - Updated backend configuration to reflect changes in reasoning model naming and supported models. - Improved agent handling logic in the orchestration layer to prevent infinite loops during approval processes.

…emove premature satisfaction guard logging in human_approval_manager.py

feat: upgrade to GPT-5.4 models, add default teams with per-user isolation, and ACR image build scripts

chore: Update Owners ID for Vinay and Prajwal

Copilot

Pull request overview

This pull request downmerges dev-v4 into main, primarily updating the accelerator’s default model configuration to GPT-5.4 (and related naming/capacity settings), while also improving post-deploy guidance and aligning infra/workflows/scripts with the new deployment model and container build flow.

Changes:

Update default model/deployment names (GPT-5.4 / GPT-5.4-mini) across backend config, agent team configs, docs, and quota/deploy workflows.
Adjust Cosmos DB team visibility and database factory behavior (default teams + per-request user scoping), plus related tests.
Improve deployment experience and infra: add ACR provisioning + post-provision image build/push scripts; update Dockerfiles to build without BuildKit mounts; optimize frontend streaming UI handling.

Reviewed changes

Copilot reviewed 44 out of 46 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
TRANSPARENCY_FAQS.md	Updates Responsible AI FAQ to reflect GPT-5.4 default LLM messaging.
src/tests/backend/common/database/test_database_factory.py	Updates tests for shared-connection + per-request DB factory semantics.
src/tests/backend/common/database/test_cosmosdb.py	Updates tests for new team query filters and delete semantics.
src/tests/agents/test_foundry_integration.py	Updates test model deployment name to GPT-5.4.
src/tests/agents/interactive_test_harness/foundry_agent_interactive.py	Updates interactive harness default model deployment name to GPT-5.4.
src/backend/v4/orchestration/human_approval_manager.py	Adds bounded “premature satisfaction” guard and normalizes agent-name matching.
src/backend/v4/common/services/team_service.py	Parses/persists `is_default` for team configs.
src/backend/Dockerfile	Removes BuildKit-only mounts; improves layer caching; ensures classic builder compatibility.
src/backend/common/models/messages_af.py	Adds `is_default` field to `TeamConfiguration`.
src/backend/common/database/database_factory.py	Switches to shared Cosmos connection + per-request user-scoped client instances.
src/backend/common/database/cosmosdb.py	Filters teams by `(user_id OR is_default)`; prevents deletion of default teams; fixes delete return semantics.
src/backend/common/config/app_config.py	Updates default deployment/model names for GPT-5.4(-mini) and reasoning model defaults.
src/backend/auth/sample_user.py	Removes sample Cookie header value.
src/backend/.env.sample	Updates sample env defaults for GPT-5.4 naming and supported models list.
src/App/src/hooks/usePlanWebSocket.tsx	Coalesces streaming tokens via `requestAnimationFrame`; ensures spinners clear on terminal statuses.
src/App/Dockerfile	Removes BuildKit-only mounts to support classic builders (e.g., `az acr build`).
infra/scripts/quota_check_params.sh	Updates default model/capacity pairs to GPT-5.4 / GPT-5.4-mini.
infra/scripts/checkquota.sh	Renames quota env vars and checks GPT-5.4 / GPT-5.4-mini capacities.
infra/scripts/Build-And-Push-Images.ps1	Adds PowerShell post-provision build/push/update script for backend/frontend/MCP images.
infra/scripts/build_and_push_images.sh	Adds Bash post-provision build/push/update script for backend/frontend/MCP images.
infra/main.waf.parameters.json	Renames/reshapes GPT model parameters to GPT-5.4 naming (incl. deployment alias).
infra/main.parameters.json	Renames/reshapes GPT model parameters to GPT-5.4 naming (incl. deployment alias).
infra/main.bicep	Updates model deployments/params, adds ACR + private DNS zone, uses placeholder images and MI-based pulls, updates outputs.
infra/main_custom.bicep	Mirrors GPT-5.4 deployment parameter/model changes for the custom template variant.
docs/TroubleShootingSteps.md	Simplifies guidance for immutable VM properties redeploy scenario.
docs/quota_check.md	Updates quota docs/examples for GPT-5.4 naming and capacities.
docs/ManualAzureDeployment.md	Updates manual deployment env example to GPT-5.4-mini deployment name.
docs/DeploymentGuide.md	Expands post-deploy steps (explicit image build/push + configuration) and updates quota/model guidance.
docs/CustomizingAzdParameters.md	Updates documented `azd` parameters for GPT-5.4 models and deployment aliases.
docs/AzureGPTQuotaSettings.md	Updates quota UI guidance to GPT-5.4 / GPT-5.4-mini.
docs/ACRBuildAndPushGuide.md	Updates build/push Dockerfile paths and clarifies post-`azd up` image workflow.
data/agent_teams/rfp_analysis_team.json	Updates default deployment_name to gpt-5.4-mini and marks as default team.
data/agent_teams/retail.json	Updates deployment_name to gpt-5.4-mini; marks as default team; aligns reasoning agent model.
data/agent_teams/marketing.json	Updates deployment_name to gpt-5.4-mini and marks as default team.
data/agent_teams/hr.json	Updates deployment_name to gpt-5.4-mini and marks as default team.
data/agent_teams/contract_compliance_team.json	Updates deployment_name to gpt-5.4-mini and marks as default team.
azure.yaml	Enhances postdeploy output with explicit build/push instructions for Bash and PowerShell.
.github/workflows/job-send-notification.yml	Renames quota env var(s) to GPT-5.4 naming.
.github/workflows/job-deploy.yml	Renames quota env var(s) and wires them into quota-check step.
.github/workflows/deploy.yml	Renames quota env var(s) and updates deploy parameters (capacity default).
.github/workflows/deploy-waf.yml	Renames quota env var(s) and updates WAF deploy parameter names.
.github/workflows/create-release.yml	Removes semantic-release-based create-release workflow.
.github/workflows/azure-dev.yml	Updates validation env vars to GPT-5.4 naming.
.github/workflows/azd-template-validation.yml	Updates validation env vars to GPT-5.4 naming.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

 ## What operational factors and settings allow for effective and responsible use of Multi Agent: Custom Automation Engine – Solution Accelerator? 
 Effective and responsible use of the Multi Agent: Custom Automation Engine – Solution Accelerator depends on several operational factors and settings. The system is designed to perform reliably and safely across a range of business tasks that it was evaluated for. Users can customize certain settings, such as the planning language model used by the system, the types of tasks that agents are assigned, and the specific actions that agents can take (e.g., sending emails or scheduling orientation sessions for new employees). However, it's important to note that these choices may impact the system's behavior in real-world scenarios.
-For example, selecting a planning language model that is not well-suited to the complexity of the tasks may result in lower accuracy and performance. Similarly, assigning tasks that are outside the system's intended scope may lead to errors or incomplete results. Users can choose the LLM that is optimized for responsible use. The default LLM is GPT-4o which inherits the existing RAI mechanisms and filters from the LLM provider. Caching is enabled by default to increase reliability and control cost. We encourage developers to review [OpenAI’s Usage policies](https://openai.com/policies/usage-policies/) and [Azure OpenAI’s Code of Conduct](https://learn.microsoft.com/en-us/legal/cognitive-services/openai/code-of-conduct) when using GPT-40. To ensure effective and responsible use of the accelerator, users should carefully consider their choices and use the system within its intended scope.
+For example, selecting a planning language model that is not well-suited to the complexity of the tasks may result in lower accuracy and performance. Similarly, assigning tasks that are outside the system's intended scope may lead to errors or incomplete results. Users can choose the LLM that is optimized for responsible use. The default LLM is GPT-5.4 which inherits the existing RAI mechanisms and filters from the LLM provider. Caching is enabled by default to increase reliability and control cost. We encourage developers to review [OpenAI’s Usage policies](https://openai.com/policies/usage-policies/) and [Azure OpenAI’s Code of Conduct](https://learn.microsoft.com/en-us/legal/cognitive-services/openai/code-of-conduct) when using GPT-5.4. To ensure effective and responsible use of the accelerator, users should carefully consider their choices and use the system within its intended scope.


+        # Ensure the shared connection infrastructure is initialized
+        if force_new or DatabaseFactory._shared_instance is None:
+            shared = CosmosDBClient(
                endpoint=config.COSMOSDB_ENDPOINT,
                credential=config.get_azure_credentials(),


 // ========== Search Service ========== //

 var searchServiceName = 'srch-${solutionSuffix}'
+var aiSearchIndexName = 'sample-dataset-index'


+        //image: '${backendContainerRegistryHostname}/${backendContainerImageName}:${backendContainerImageTag}'
+        image: 'mcr.microsoft.com/azuredocs/containerapps-helloworld:latest'


…ment-macae fix: remove local deployment related files

Copilot

Pull request overview

Copilot reviewed 47 out of 49 changed files in this pull request and generated 3 comments.

Comments suppressed due to low confidence (1)

infra/main.bicep:1257

Operational: infra/main.bicep now hard-codes placeholder containerapps-helloworld images for both backend and MCP. This makes deployments non-functional unless a separate post-step updates the images, and it also means the existing image parameters (backendContainer*, MCPContainer*, etc.) no longer affect the deployed services. Consider adding a parameterized switch (e.g., usePlaceholderImages) so CI/prod can deploy real images while first-time azd up can still succeed without prebuilt images.

        name: 'backend'
        //image: '${backendContainerRegistryHostname}/${backendContainerImageName}:${backendContainerImageTag}'
        image: 'mcr.microsoft.com/azuredocs/containerapps-helloworld:latest'
        resources: {

                plan=json_data.get("plan", ""),
                starting_tasks=starting_tasks,
                user_id=user_id,
+                is_default=json_data.get("is_default", False),
            )


              gptModelVersion="2025-04-14" \
              backendContainerImageTag="${IMAGE_TAG}" \
              frontendContainerImageTag="${IMAGE_TAG}" \
              azureAiServiceLocation='${{ env.AZURE_LOCATION }}' \
-              gptDeploymentCapacity=50 \
+              gptDeploymentCapacity=100 \


+if ($deploymentType -eq 'WAF') {
+    Write-Section 'WAF deployment detected - temporarily relaxing ACR restrictions'
+    & az acr update --name $acrName --resource-group $resourceGroup --allow-exports true --output none
+    if ($LASTEXITCODE -ne 0) { throw "Failed to enable ACR exports." }
+    & az acr update --name $acrName --resource-group $resourceGroup --public-network-enabled true --output none
+    if ($LASTEXITCODE -ne 0) { throw "Failed to enable ACR public network access." }
+    & az acr update --name $acrName --resource-group $resourceGroup --default-action Allow --output none
+    if ($LASTEXITCODE -ne 0) { throw "Failed to set ACR default action to Allow." }
+    Write-Host 'ACR restrictions temporarily relaxed.'
+}


Roopan-Microsoft and others added 25 commits May 25, 2026 20:50

Merge pull request #1001 from microsoft/dev-v4

22aa354

fix: Dev v4 to main PR

fix: Dev v4 to main merge

bf9162d

deleting the create-release.yml (codfish) due to security issue

7b76ce1

Merge pull request #1069 from AbdulMujeeb-2000/main

b4a4a00

chore: delete Create-Release workflow due to codfish security incident

chore: update CODEOWNERS handles

1e155a1

Replace @Vinay-Microsoft -> @VinaySh-Microsoft and @Prajwal-Microsoft -> @Prajwal1-Microsoft

chore: update CODEOWNERS handles

4265284

updated the models and configs

4ec711f

updated model to gpt5.4 models

f1b191e

fix: remove sensitive cookie data from sample user configuration

cb64ca9

Merge pull request #1083 from microsoft/macae-ayaz

f581a71

fix: remove sensitive cookie data from sample user configuration

Updated team config with isdefault and logic to display teams is updated

9fd3e4d

Merge remote-tracking branch 'origin/main' into psl-mainChanges

440824b

update deployment guide and azure.yaml

21fc8d1

fix test cases

56c0411

updated model name related docs and var

663f350

updated var names according to gpt model

4541d7d

update environment variable names for GPT models in workflows and scr…

584bcbc

…ipts

add instructions for uploading team configurations and indexing sampl…

212f7c5

…e data after image build and push

remove unused instances list from TestDatabaseFactoryGetDatabase

1ab8138

update minimum required capacities for GPT models in checkquota.sh; r…

91e3317

…emove premature satisfaction guard logging in human_approval_manager.py

Merge pull request #1082 from microsoft/psl-mainChanges

452419c

feat: upgrade to GPT-5.4 models, add default teams with per-user isolation, and ACR image build scripts

Update Owners ID for Vinay and Prajwal

f323d53

Update Owners ID for Vinay and Prajwal 2

a115644

Merge pull request #1093 from chaudhariniraj/main

abd2781

chore: Update Owners ID for Vinay and Prajwal

Copilot AI review requested due to automatic review settings July 17, 2026 04:32

Dhruvkumar-Microsoft requested review from Avijit-Microsoft, Fr4nc3, Roopan-Microsoft and aniaroramsft as code owners July 17, 2026 04:32

Dhruvkumar-Microsoft requested review from dgp10801, nchandhi and toherman-msft as code owners July 17, 2026 04:32

Copilot started reviewing on behalf of Dhruvkumar-Microsoft July 17, 2026 04:32 View session

Copilot AI reviewed Jul 17, 2026

View reviewed changes

Priyanka2-Microsoft and others added 2 commits July 17, 2026 13:56

remove local deployment related files

c1869ab

Merge pull request #1101 from Priyanka2-Microsoft/remove-local-deploy…

12262e2

…ment-macae fix: remove local deployment related files

Copilot AI review requested due to automatic review settings July 20, 2026 06:52

Roopan-Microsoft requested review from Prajwal-Microsoft and Vinay-Microsoft as code owners July 20, 2026 06:52

Copilot started reviewing on behalf of Roopan-Microsoft July 20, 2026 06:52 View session

Copilot AI reviewed Jul 20, 2026

View reviewed changes

Roopan-Microsoft closed this Jul 20, 2026

Roopan-Microsoft had a problem deploying to production July 23, 2026 02:17 — with GitHub Actions Failure

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

fix: dev-v4 to main branch downmerge - #1098

fix: dev-v4 to main branch downmerge#1098
Dhruvkumar-Microsoft wants to merge 27 commits into
dev-v4from
main

Dhruvkumar-Microsoft commented Jul 17, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

11 participants

		//image: '${backendContainerRegistryHostname}/${backendContainerImageName}:${backendContainerImageTag}'
		image: 'mcr.microsoft.com/azuredocs/containerapps-helloworld:latest'

Uh oh!

Conversation

Dhruvkumar-Microsoft commented Jul 17, 2026

Purpose

Does this introduce a breaking change?

How to Test

What to Check

Other Information

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

11 participants