Skip to content

Delete Cap Host Details#166

Closed
meerakurup wants to merge 243 commits intomicrosoft-foundry:mainfrom
meerakurup:main
Closed

Delete Cap Host Details#166
meerakurup wants to merge 243 commits intomicrosoft-foundry:mainfrom
meerakurup:main

Conversation

@meerakurup
Copy link
Copy Markdown
Contributor

Deleting the cap host file added and README updated

foundry-samples-repo-sync bot and others added 26 commits February 23, 2026 16:59
Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
* changed bicep params and added testing guide

* updated testing guide

* removed sub id from test guide

* MCP server deployed as ACA and not as ACI

* added mcp subnet

* fixes in main.bicep and network-agent-vnet.bicep

* updated testing guide to account for the MCP server container app

* Revert template 15 to upstream and add template 19 (hybrid private resources)

- Reverted 15-private-network-standard-agent-setup to match upstream/main exactly
- Added 19-hybrid-private-resources-agent-setup: hybrid architecture with public
  AI Services endpoint and private backend resources (AI Search, Cosmos DB, Storage)
- Template 19 enables portal-based agent testing while keeping data resources private
- Includes TESTING-GUIDE.md, test scripts, and architecture diagrams in template 19

* Update template 19 testing guide and add MCP HTTP server

Updates based on actual testing:
- Fixed SDK usage: use AzureAISearchTool class, not dict format
- Added note about portal limitation with network injection
- Added correct API version (2025-05-15-preview) for REST API
- Added MCP HTTP server implementation (Streamable HTTP transport)
- Updated MCP deployment instructions with proper HTTP-based server
- Added connection name lookup command
- Fixed test script to use AzureAISearchTool class

MCP Server:
- Added mcp-http-server/server.py - Flask-based MCP server
- Added mcp-http-server/Dockerfile for containerization
- Implements JSON-RPC over HTTP as required by Azure AI Agents

* updated mcp server and tests

* added PNA disabled for Foundry in bicep + test cleanup

* Automatic fixes

---------

Co-authored-by: Meera Kurup <meerakurup@microsoft.com>
Co-authored-by: foundry-samples automation <foundry-samples@noreply.github.com>
When deploying AI projects with connections (CosmosDB, Azure Storage, AI Search),
the deployment was failing with a 409 Conflict error because all 3 connection
resources were being created in parallel by default, and the backend service
doesn't support concurrent connection creation on the same project.

This change serializes the connection creation so they deploy one after another:
1. CosmosDB connection deploys first (no dependencies)
2. Azure Storage connection waits for CosmosDB connection to complete
3. AI Search connection waits for Azure Storage connection to complete

Changes:
- ai-project-identity.bicep: Added dependsOn chains between connection resources
- azuredeploy.json: Added corresponding dependencies in ARM template

Co-authored-by: jiyan2 <jiyan@microsoft.com>
Co-authored-by: jblaaa-codes <jbla9028@gmail.com>
ARM templates do not support 'comment' property on resource objects.
This was causing InvalidRequestContent error during deployment.
The dependsOn fixes from PR #560 are preserved.

Co-authored-by: jiyan2 <jiyan@microsoft.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
…t, also adding samples folder and adding missing pdf/docx there. (#564)
Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
* adding fabric PE to the vnet bicep template

* changed bicep params and added testing guide

* updated testing guide

* removed sub id from test guide

* MCP server deployed as ACA and not as ACI

* added mcp subnet

* fixes in main.bicep and network-agent-vnet.bicep

* updated testing guide to account for the MCP server container app

* Revert template 15 to upstream and add template 19 (hybrid private resources)

- Reverted 15-private-network-standard-agent-setup to match upstream/main exactly
- Added 19-hybrid-private-resources-agent-setup: hybrid architecture with public
  AI Services endpoint and private backend resources (AI Search, Cosmos DB, Storage)
- Template 19 enables portal-based agent testing while keeping data resources private
- Includes TESTING-GUIDE.md, test scripts, and architecture diagrams in template 19

* Update template 19 testing guide and add MCP HTTP server

Updates based on actual testing:
- Fixed SDK usage: use AzureAISearchTool class, not dict format
- Added note about portal limitation with network injection
- Added correct API version (2025-05-15-preview) for REST API
- Added MCP HTTP server implementation (Streamable HTTP transport)
- Updated MCP deployment instructions with proper HTTP-based server
- Added connection name lookup command
- Fixed test script to use AzureAISearchTool class

MCP Server:
- Added mcp-http-server/server.py - Flask-based MCP server
- Added mcp-http-server/Dockerfile for containerization
- Implements JSON-RPC over HTTP as required by Azure AI Agents

* updated mcp server and tests

* added PNA disabled for Foundry in bicep + test cleanup

* Automatic fixes

---------

Co-authored-by: Meera Kurup <meerakurup@microsoft.com>
Co-authored-by: foundry-samples automation <foundry-samples@noreply.github.com>
* Redirect contributors to foundry-samples-pr staging repo

Rewrite CONTRIBUTING.md to explain the private-to-public workflow:
- External users: open issues for bugs/suggestions
- Microsoft contributors: join microsoft-foundry org, then contribute
  via the private foundry-samples-pr staging repo

Update README.md contributing section to match.

The previous fork-based workflow and stale azure-ai-foundry/doc-samples
references are removed.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Update contributing guidelines for Microsoft Foundry

* Add workflow to auto-close PRs and redirect contributors

Automatically closes PRs with a comment directing:
- Microsoft org members to foundry-samples-pr
- External contributors to open an issue instead

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Add explicit permissions and duplicate comment prevention

- Declare pull-requests: write permission (principle of least privilege)
- Check for existing bot comments before posting (idempotent on re-runs)

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

---------

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Move each quickstart sample into its own subdirectory (chat-with-agent,
create-agent, responses) with per-sample requirements.txt and sample.yaml
files. This aligns with the foundry-samples-pr repo structure.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Add a named bot allowlist to the redirect-pull-requests workflow so
that PRs created by foundry-samples-repo-sync[bot] are not auto-closed.
The nightly sync from foundry-samples-pr needs these PRs to stay open
for review and merge.

Fixes ADO#5063962

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* updating quickstart samples

* variable fixes

* updating java

* fixes

---------

Co-authored-by: Aaron Hill <aahi@microsoft.com>
…re, prefix removal) (#580)

* linting

* Update .gitignore to modify ignored files

Removed unnecessary entries from .gitignore and added back the developer-journey-stage-1-idea-to-prototype.md file.

* removing prefix

---------

Co-authored-by: Aaron Hill <aahi@microsoft.com>
Co-authored-by: aahill <aahill017@live.com>
Replace the single 'microsoft' org membership check with a multi-signal
cascade that catches internal contributors with private org membership:

1. Check microsoft-foundry org membership (full GITHUB_TOKEN visibility)
2. Check repo collaborator status
3. Check microsoft org membership (public members fallback)

Also adds members: read permission and debug logging.

Fixes misidentification of internal contributors like #574 where a
Microsoft docs writer was incorrectly treated as an external contributor.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
The 'members: read' permission key added in #586 is not a valid
GitHub Actions permission, causing the workflow YAML to fail
validation. This broke auto-closing of PRs (#587, #588).

The repos.checkCollaborator API works with the default GITHUB_TOKEN
repo access — no extra permission is needed.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* Sync from private repository (bootstrap)

* Record sync state

---------

Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
* Restore: 19-hybrid-private-resources-agent-setup (content lost to repo sync) (#77)

* Restore 19-hybrid-private-resources-agent-setup (lost to repo sync)

Add the full 19-hybrid-private-resources-agent-setup bicep template to the
staging repo. This content was contributed by @meerakurup and collaborators
directly to the public foundry-samples repo, but was never added to this
staging repo. As a result, the nightly sync automation repeatedly deleted
it from the public repo (most recently via PRs #569 on 2026-03-02 and
#587 on 2026-03-10).

Content restored from foundry-samples commit 99a8388 (the last commit
before the most recent sync wipe).

Files restored (42 total):
- Bicep templates: main.bicep, add-project.bicep, azuredeploy.json, etc.
- Modules: 18 bicep modules in modules-network-secured/
- Scripts: createCapHost.sh, deleteCapHost.sh, get-existing-resources.ps1
- Diagrams: 5 architecture PNGs + generator script
- Tests: 3 Python test files + testing guide
- Config: metadata.json, parameter files, Dockerfile

Co-authored-by: Meera Kurup <meerakurup@microsoft.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Automatic fixes

---------

Co-authored-by: Meera Kurup <meerakurup@microsoft.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: foundry-samples automation <foundry-samples@noreply.github.com>

* Update sync state to c0591586a2f2

---------

Co-authored-by: Brandon Miller <42854725+brandom-msft@users.noreply.github.com>
Co-authored-by: Meera Kurup <meerakurup@microsoft.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: foundry-samples automation <foundry-samples@noreply.github.com>
Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
* Update agent initialization and requirements for Azure AI Agent Framework (#79)

* Update sync state to f65b829bf46c

---------

Co-authored-by: Hui Miao <miuhui@outlook.com>
Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
* initial commit (#83)

* updating C# code (#90)

* Update sync state to 560fc8afc9b5

---------

Co-authored-by: nikhowlettMicrosoft <97988752+nikhowlettMicrosoft@users.noreply.github.com>
Co-authored-by: aahill <aahi@microsoft.com>
Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
* fix: scope Cosmos DB SQL role assignment to account level for Agents v2 (#84)

Replace three per-collection azurerm_cosmosdb_sql_role_assignment resources
(scoped to individual enterprise_memory collections) with a single assignment
scoped at the Cosmos DB account level. This is required for Agents v2 and
eliminates redundant per-collection definitions.

Affected templates:
- infrastructure-setup-terraform/15a-private-network-standard-agent-setup
- infrastructure-setup-terraform/15b-private-network-standard-agent-setup-byovnet

* Update README with correct parameters for deployment (#95)

* fixup: correct drift after commit replay

* Update sync state to 4144addd759b

---------

Co-authored-by: geabdluca <73857153+geabdluca@users.noreply.github.com>
Co-authored-by: AhmadAbdullah91 <91725950+AhmadAbdullah91@users.noreply.github.com>
Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
* Updating to new java packages (#93)

* updating C# code

* updating python variables

* updating example

* updating rest samples

* Update sync state to bfdb8b40dbde

---------

Co-authored-by: aahill <aahi@microsoft.com>
Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
* Add Foundry Local code samples for Python, C#, JavaScript, and Rust (#85)

* Add Foundry Local code samples for Python, C#, JavaScript, and Rust

Extract application code samples from azure-ai-docs-pr Foundry Local documentation
into standalone, runnable sample files organized by language and scenario.

Languages and scenarios:
- Python: integrate-sdks, langchain-integration, compile-hugging-face-models
- C#: integrate-sdks, audio-transcription, native-chat-completions, tool-calling-native, tool-calling-openai
- JavaScript: integrate-sdks, audio-transcription, native-chat-completions, tool-calling, langchain-integration
- Rust: integrate-sdks

Each sample includes:
- Section tags for docs-build inclusion (# <name> / # </name>)
- Project/dependency files (requirements.txt, package.json, .csproj, Cargo.toml)
- Self-contained code (C# Utils references replaced with inline implementations)

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Fix Betalgo.Ranul.OpenAI version to 9.2.6 in C# projects

Version 9.5.0 does not exist on NuGet. The latest available is 9.2.6.

Note: native-chat-completions and tool-calling-native projects that use
the Betalgo-based native ChatClient API have a runtime incompatibility
with Foundry.Local SDK 0.9.0 (ReasoningEfforts type not found). This is
a pre-existing SDK compatibility issue, not a sample code issue.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Fix Rust sample: bypass SDK catalog parsing, use REST directly

The foundry-local Rust crate v0.1.0 fails to deserialize
NvTensorRTRTXExecutionProvider from the Foundry Local v0.8.119 catalog.
Rewrote to use reqwest directly against the REST API, which works
reliably. Also fixed Betalgo.Ranul.OpenAI version to 9.2.6.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

---------

Co-authored-by: Jon Burchel <jonburchel@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Update sync state to a62fc7d87bb4

---------

Co-authored-by: Jon Burchel <jburchel@microsoft.com>
Co-authored-by: Jon Burchel <jonburchel@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
* tf samples to match bicep (#101)

* ci: add CODEOWNERS entries (#106)

* Update sync state to 12648bcde39f

---------

Co-authored-by: Dennis Eikelenboom <92387505+deeikele@users.noreply.github.com>
Co-authored-by: Hui Miao <miuhui@outlook.com>
Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
* Connection Migration for Tools (Grounding, AI Search, Fabric) (#103)

* initial commit

* save derived arm name

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* use display name in connection map

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* connection fixes

* always update display ame

* update source->target dictionary

* update connection map language

* use correct token

---------

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* add provision network now and chain dependency in tf for connections (#133)

* Updates for Java SDK GA (#135)

* Update sync state to 8dc749def5ba

---------

Co-authored-by: nikhowlettMicrosoft <97988752+nikhowlettMicrosoft@users.noreply.github.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Dchillakuru <167816084+Dchillakuru@users.noreply.github.com>
Co-authored-by: Alvin Ashcraft <73072+alvinashcraft@users.noreply.github.com>
Co-authored-by: foundry-samples-repo-sync[bot] <foundry-samples-repo-sync[bot]@users.noreply.github.com>
@meerakurup meerakurup requested review from a team as code owners March 31, 2026 21:32
@meerakurup meerakurup closed this by deleting the head repository Mar 31, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.