Python: Add more types - #5377
Closed
Tao Chen (TaoChenOSU) wants to merge 13 commits into
Closed
Conversation
* Experiment * Update dependency and add non streaming * Add more samples * Rename samples * Add invocations * Comments 1 * Comments 2 * Comments 3 * Improve README * Add local shell sample * WIP: Add eval and memory samples * Update user agent prefix * Update user agent prefix doc
* Add tests * fix tests and sample * Fix formatting * Remove function approval contents
* Refine samples and upgrade pacakges * Upgrade to a new package that fixes a bug * Update model env var
* Upgrade agentserver packages * Fix new types
* Add special handling for workflows * Address comments
Evan Mattson (moonbox3)
added
documentation
Contributor
Python Test Coverage Report •
Python Unit Test Overview
|
|||||||||||||||||||||||||||||||||||||||||||||
Contributor
There was a problem hiding this comment.
Pull request overview
This PR expands the Python side of the Agent Framework repo by adding a new Foundry Hosting integration package, updating core telemetry User-Agent handling, and replacing older “hosted agent” end-to-end samples with a new set of Foundry Hosted Agents samples (Responses + Invocations).
Changes:
- Add new
agent-framework-foundry-hostingpackage (ResponsesHostServer / InvocationsHostServer) plus HTTP round-trip tests for Responses. - Update core telemetry to support adding hosting-layer prefixes to the framework User-Agent.
- Restructure hosting samples: remove older
05-end-to-end/hosted_agents/*samples and add new04-hosting/foundry-hosted-agents/*sample set (code + manifests + Docker assets).
Reviewed changes
Copilot reviewed 81 out of 85 changed files in this pull request and generated 29 comments.
Show a summary per file
| File | Description |
|---|---|
| python/uv.lock | Adds workspace member and locks new Foundry Hosting/Azure AgentServer dependencies. |
| python/pyproject.toml | Registers agent-framework-foundry-hosting as a workspace package. |
| python/.cspell.json | Adds agentserver to spellchecker dictionary. |
| python/packages/core/agent_framework/_telemetry.py | Adds support for User-Agent prefix stacking for hosting layers. |
| python/packages/foundry_hosting/pyproject.toml | New package metadata, dependencies, and test/tooling configuration. |
| python/packages/foundry_hosting/agent_framework_foundry_hosting/init.py | Exposes ResponsesHostServer and InvocationsHostServer. |
| python/packages/foundry_hosting/agent_framework_foundry_hosting/_responses.py | Implements Responses API hosting bridge and content/event conversions. |
| python/packages/foundry_hosting/agent_framework_foundry_hosting/_invocations.py | Implements Invocations API hosting bridge. |
| python/packages/foundry_hosting/tests/test_responses.py | Adds comprehensive HTTP round-trip tests for ResponsesHostServer. |
| python/packages/foundry_hosting/README.md | Introduces package README (currently placeholder content). |
| python/packages/foundry_hosting/LICENSE | Adds MIT license for the new package. |
| python/samples/04-hosting/foundry-hosted-agents/README.md | New top-level README for Foundry Hosted Agents samples. |
| python/samples/04-hosting/foundry-hosted-agents/responses/README.md | Adds Responses API sample index + “using_deployed_agent.py” helper script. |
| python/samples/04-hosting/foundry-hosted-agents/responses/using_deployed_agent.py | Demonstrates calling a deployed agent via OpenAIChatClient. |
| python/samples/04-hosting/foundry-hosted-agents/responses/01_basic/requirements.txt | Sample requirements for basic Responses-hosted agent. |
| python/samples/04-hosting/foundry-hosted-agents/responses/01_basic/main.py | Basic Responses-hosted agent server sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/01_basic/agent.yaml | Hosted agent definition for Responses basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/01_basic/agent.manifest.yaml | Deployable manifest for Responses basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/01_basic/README.md | How-to for the Responses basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/01_basic/Dockerfile | Container build for the Responses basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/01_basic/.env.example | Example env vars for the Responses basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/01_basic/.dockerignore | Build context excludes for the Responses basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/02_local_tools/requirements.txt | Sample requirements for Responses + local tools. |
| python/samples/04-hosting/foundry-hosted-agents/responses/02_local_tools/main.py | Responses-hosted agent demonstrating local tools (incl. shell tool). |
| python/samples/04-hosting/foundry-hosted-agents/responses/02_local_tools/agent.yaml | Hosted agent definition for local tools sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/02_local_tools/agent.manifest.yaml | Deployable manifest for local tools sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/02_local_tools/README.md | How-to for local tools sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/02_local_tools/Dockerfile | Container build for local tools sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/02_local_tools/.env.example | Example env vars for local tools sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/02_local_tools/.dockerignore | Build context excludes for local tools sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/03_remote_mcp/requirements.txt | Sample requirements for Responses + remote MCP tools. |
| python/samples/04-hosting/foundry-hosted-agents/responses/03_remote_mcp/main.py | Responses-hosted agent demonstrating Foundry toolbox MCP + GitHub MCP. |
| python/samples/04-hosting/foundry-hosted-agents/responses/03_remote_mcp/agent.yaml | Hosted agent definition for remote MCP sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/03_remote_mcp/agent.manifest.yaml | Deployable manifest for remote MCP sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/03_remote_mcp/README.md | How-to for remote MCP sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/03_remote_mcp/Dockerfile | Container build for remote MCP sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/03_remote_mcp/.env.example | Example env vars for remote MCP sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/03_remote_mcp/.dockerignore | Build context excludes for remote MCP sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/04_workflows/requirements.txt | Sample requirements for Responses-hosted workflow. |
| python/samples/04-hosting/foundry-hosted-agents/responses/04_workflows/main.py | Responses-hosted workflow sample using Agent Framework workflows. |
| python/samples/04-hosting/foundry-hosted-agents/responses/04_workflows/agent.yaml | Hosted agent definition for workflow sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/04_workflows/agent.manifest.yaml | Deployable manifest for workflow sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/04_workflows/README.md | How-to for workflow sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/04_workflows/Dockerfile | Container build for workflow sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/04_workflows/.env.example | Example env vars for workflow sample. |
| python/samples/04-hosting/foundry-hosted-agents/responses/04_workflows/.dockerignore | Build context excludes for workflow sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/README.md | Adds Invocations API sample index. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/01_basic/requirements.txt | Sample requirements for basic Invocations-hosted agent. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/01_basic/main.py | Basic Invocations-hosted agent server sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/01_basic/agent.yaml | Hosted agent definition for Invocations basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/01_basic/agent.manifest.yaml | Deployable manifest for Invocations basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/01_basic/README.md | How-to for Invocations basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/01_basic/Dockerfile | Container build for Invocations basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/01_basic/.env.example | Example env vars for Invocations basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/01_basic/.dockerignore | Build context excludes for Invocations basic sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/02_break_glass/requirements.txt | Sample requirements for Invocations “break glass” scenario. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/02_break_glass/main.py | Invocations sample with custom invoke handler and session store. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/02_break_glass/agent.yaml | Hosted agent definition for “break glass” sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/02_break_glass/agent.manifest.yaml | Deployable manifest for “break glass” sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/02_break_glass/README.md | How-to for “break glass” sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/02_break_glass/Dockerfile | Container build for “break glass” sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/02_break_glass/.env.example | Example env vars for “break glass” sample. |
| python/samples/04-hosting/foundry-hosted-agents/invocations/02_break_glass/.dockerignore | Build context excludes for “break glass” sample. |
| python/samples/05-end-to-end/hosted_agents/README.md | Removes older hosted agent sample index under 05-end-to-end. |
| python/samples/05-end-to-end/hosted_agents/agents_in_workflow/requirements.txt | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agents_in_workflow/main.py | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agents_in_workflow/agent.yaml | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_text_search_rag/requirements.txt | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_text_search_rag/main.py | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_text_search_rag/agent.yaml | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_local_tools/requirements.txt | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_local_tools/main.py | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_local_tools/agent.yaml | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_local_tools/README.md | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_local_tools/.env.sample | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_local_tools/.dockerignore | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_hosted_mcp/requirements.txt | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_hosted_mcp/main.py | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/agent_with_hosted_mcp/agent.yaml | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/writer_reviewer_agents_in_workflow/requirements.txt | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/writer_reviewer_agents_in_workflow/main.py | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/writer_reviewer_agents_in_workflow/agent.yaml | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/writer_reviewer_agents_in_workflow/README.md | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/writer_reviewer_agents_in_workflow/.env.sample | Removes older end-to-end hosted agent sample assets. |
| python/samples/05-end-to-end/hosted_agents/writer_reviewer_agents_in_workflow/.dockerignore | Removes older end-to-end hosted agent sample assets. |
Comment on lines
+21
to
+22
| Send a POST request to the server with a JSON body containing a "message" field to interact with the agent. For example: | ||
|
|
There was a problem hiding this comment.
This README says requests should include a JSON body with a "message" field, but the Responses API (and the curl examples below) uses the "input" field. Please update the text to avoid confusing users.
Comment on lines
+33
to
+43
| # Foundry Toolbox as a MCP tool | ||
| project_endpoint = os.environ["FOUNDRY_PROJECT_ENDPOINT"] | ||
| toolbox_name = os.environ["TOOLBOX_NAME"] | ||
| toolbox_endpoint = f"{project_endpoint.rstrip('/')}/toolboxes/{toolbox_name}/mcp?api-version=v1" | ||
| http_client = httpx.AsyncClient(auth=ToolboxAuth(), headers={"Foundry-Features": "Toolboxes=V1Preview"}) | ||
| foundry_mcp_tool = MCPStreamableHTTPTool( | ||
| name="toolbox", | ||
| url=toolbox_endpoint, | ||
| http_client=http_client, | ||
| load_prompts=False, | ||
| ) |
There was a problem hiding this comment.
http_client = httpx.AsyncClient(...) is created but never closed. Since this sample runs as a long-lived server, this can leak connections/resources; consider using a lifespan/shutdown hook to aclose() the client or manage it with an async context manager.
| @@ -0,0 +1,2 @@ | |||
| FOUNDRY_PROJECT_ENDPOINT= "..." | |||
There was a problem hiding this comment.
There is an extra space after = in FOUNDRY_PROJECT_ENDPOINT= "...", which can result in the value containing a leading space depending on the dotenv parser. Please remove the extra space (and keep quoting consistent with the other examples).
Suggested change
| FOUNDRY_PROJECT_ENDPOINT= "..." | |
| FOUNDRY_PROJECT_ENDPOINT="..." |
Comment on lines
+35
to
+36
| query = "You name is Javis. What can you do?" | ||
| print(f"\nUser: {query}") |
There was a problem hiding this comment.
The sample prompt has a grammatical typo: "You name is ..." should be "Your name is ..." (and confirm whether the intended name is "Jarvis"). This is user-facing sample text and will be copied verbatim by readers.
Comment on lines
+59
to
+69
| if stream: | ||
|
|
||
| async def stream_response() -> AsyncGenerator[str]: | ||
| async for update in self._agent.run(user_message, session=session, stream=True): | ||
| yield update.text | ||
|
|
||
| return StreamingResponse( | ||
| stream_response(), | ||
| media_type="text/event-stream", | ||
| headers={"Cache-Control": "no-cache", "Connection": "keep-alive"}, | ||
| ) |
There was a problem hiding this comment.
The streaming path sets media_type="text/event-stream" but yields raw text chunks (update.text) without SSE framing (data: ...\n\n). If clients interpret this as SSE, they may fail to parse the stream. Consider either formatting each chunk as a proper SSE event, or change the content-type to text/plain/chunked text if SSE isn’t intended.
| ```bash | ||
| docker run -p 8088:8088 \ | ||
| -e FOUNDRY_PROJECT_ENDPOINT=<your-endpoint> \ | ||
| -e FOUNDRY_MODEL=<your-model> \ |
There was a problem hiding this comment.
The top-level README uses FOUNDRY_MODEL in the docker run example, but the samples’ code (and .env.example files) expect MODEL_DEPLOYMENT_NAME. Please align the documented env var names to what the code actually reads to avoid startup failures.
Suggested change
| -e FOUNDRY_MODEL=<your-model> \ | |
| -e MODEL_DEPLOYMENT_NAME=<your-model> \ |
Comment on lines
+19
to
+21
| Send a POST request to the server with a JSON body containing a "message" field to interact with the agent. For example: | ||
|
|
||
| ```bash |
There was a problem hiding this comment.
This README says requests should include a JSON body with a "message" field, but the Responses API (and the curl example below) uses the "input" field. Please update the text to avoid confusing users.
| | Sample | Description | | ||
| | --- | --- | | ||
| | [01_basic](./01_basic) | A basic example of hosting an agent with the `invocations` API and carrying on a multi-turn conversation. | | ||
| | [02_break_glass](./02-break-glass) | An example of hosting an agent with the `invocations` API and a "break glass" scenario where you can create your own `invoke_handler` to handle specific types of invocations. | |
There was a problem hiding this comment.
The link target uses ./02-break-glass, but the folder added in this PR is 02_break_glass. Please update the README link so it resolves correctly.
Suggested change
| | [02_break_glass](./02-break-glass) | An example of hosting an agent with the `invocations` API and a "break glass" scenario where you can create your own `invoke_handler` to handle specific types of invocations. | | |
| | [02_break_glass](./02_break_glass) | An example of hosting an agent with the `invocations` API and a "break glass" scenario where you can create your own `invoke_handler` to handle specific types of invocations. | |
Comment on lines
+1
to
+11
| # Foundry Hosting | ||
|
|
||
| This package provides the integration of Agent Framework agents and workflows with the Foundry Agent Server, which can be hosted on Foundry infrastructure. | ||
|
|
||
| ## Responses | ||
|
|
||
| TODO | ||
|
|
||
| ## Invocations | ||
|
|
||
| TODO |
There was a problem hiding this comment.
This package README still contains placeholder "TODO" sections. Since this is a new published workspace package, please replace the TODOs with at least minimal usage examples (how to host an agent via ResponsesHostServer/InvocationsHostServer) and any required environment variables.
Comment on lines
+162
to
+170
| async def _handler( | ||
| self, | ||
| request: CreateResponse, | ||
| context: ResponseContext, | ||
| cancellation_signal: asyncio.Event, | ||
| ) -> AsyncIterable[ResponseStreamEvent | dict[str, Any]]: | ||
| """Handle the creation of a response.""" | ||
| if self._is_workflow_agent: | ||
| # Workflow agents are handled differently because they require checkpoint restoration |
There was a problem hiding this comment.
cancellation_signal is accepted by _handler / _handle_workflow_agent but never checked. If the hosting layer sets this event on client disconnect/cancel, the current implementation will keep running the agent/workflow unnecessarily. Consider periodically checking cancellation_signal.is_set() (especially inside the streaming loops) and stopping early / closing builders.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Motivation and Context
Description
Contribution Checklist