Repository navigation
Add SandboxAQ Foundry model agent - #230
Closed
sravankumarux wants to merge 3 commits into
Closed
sravankumarux wants to merge 3 commits into
sravankumarux wants to merge 3 commits into
Conversation
Per review discussion, the Foundry model agent (agents/sandboxaq) moves to its own draft PR so the MCP agents (aqcat, aqpotency) and the Foundry agent are reviewed separately. Drop the sandboxaq links from the aqcat and aqpotency READMEs.
Calls a SandboxAQ model subscribed through Azure AI Foundry from a Discovery tool container (sandboxaq-model). Split out from the AQCat and AQPotency MCP agents so it can be reviewed separately; the README no longer links to those agents.
The build command ran `docker build .` from the agent directory, but the Dockerfile lives in tools/sandboxaq-model/, so the build failed. Build from the tool directory, and add --platform linux/amd64 as the other catalog agents do, since the tool runs on x86-64 Azure VMs. Also fence the architecture diagram and bracket the support email.
Contributor
Author
|
Superseded by #231, which makes the same net README changes in a single PR. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Re-proposes the SandboxAQ Foundry model agent (
agents/sandboxaq) on its own, split out from #193 as discussed with the Discovery team. The AQCat and AQPotency MCP agents are reviewed separately.Depends on #229, which removes this agent from
main. This branch is stacked on #229, so until #229 merges the diff here shows only the README change. Once it merges, the diff becomes the full agent addition.What the agent does
Invokes a SandboxAQ model that the customer has subscribed to in Azure AI Foundry, through its chat-completions endpoint, from a Discovery tool container (
sandboxaq-model). The agent writes Python scripts that callsandboxaq_utils.invoke_model, so prompt text never passes through a shell.What's included
agents/sandboxaq/:agent.yaml,metadata.yaml,README.mdagents/sandboxaq/tools/sandboxaq-model/:Dockerfile(onmcr.microsoft.com/azurelinux/base/python:3.12),tool.yaml,sandboxaq_utils.pyChange from the version merged in #193: the README no longer links to the
aqcatandaqpotencyagents.Configuration
No secrets are committed. The tool reads runtime environment variables set at deploy time, following the same pattern as the PubMed agent:
FOUNDRY_ENDPOINTFOUNDRY_API_KEYFOUNDRY_DEPLOYMENTazure-openaiFOUNDRY_API_STYLEazure-openai(default,api-keyheader) orserverless(Bearer token,/chat/completions)FOUNDRY_API_VERSION2024-10-21forazure-openaiErrors raised by the tool report the HTTP status and response body but never the endpoint URL or credential.
Status
Draft: not yet deployed end to end against a subscribed SandboxAQ Foundry deployment, and the
serverlessrequest style has not been tested against a live endpoint.