Python: fix get-started samples env wiring and FunctionalWorkflow.build() removal

[Linwen (linwendeng)]

Motivation & Context

The python/samples/01-get-started/ samples had two rough edges that blocked a
first-time user from running them out of the box:

1. Samples 02_add_tools.py, 03_multi_turn.py, and 04_memory.py hard-coded
   project_endpoint="https://your-project.services.ai.azure.com" and
   model="gpt-4o", so they crashed until the reader edited the source. Sample
   01_hello_agent.py already read from FOUNDRY_PROJECT_ENDPOINT / FOUNDRY_MODEL
   — this PR brings the other three in line with it.
2. Samples 05_functional_workflow_with_agents.py and
   06_functional_workflow_basics.py called poem_workflow.build() /
   text_workflow.build(). With the currently published agent_framework
   package, the @workflow decorator returns a ready-to-run FunctionalWorkflow
   directly, so .build() raises AttributeError.

The README in the same folder didn't cover the Foundry data-plane RBAC needed to
call the project-scoped Responses endpoint, so users landed on a 401/403 with no
in-repo pointer.

Description & Review Guide

• What are the major changes?

  • 01_hello_agent.py, 02_add_tools.py, 03_multi_turn.py, 04_memory.py:
    call load_dotenv(), read FOUNDRY_PROJECT_ENDPOINT and FOUNDRY_MODEL
    from the environment, and default the model to gpt-4.1-mini instead of
    gpt-4o.
  • 05_functional_workflow_with_agents.py,
    06_functional_workflow_basics.py: drop the obsolete
    workflow_instance = <name>.build() step and call .run() directly on the
    FunctionalWorkflow returned by the @workflow decorator.
  • README.md: adds a "Configure environment variables" section (endpoint
    shape, .env example, and a note on python-dotenv's upward search plus
    shell-env precedence) and a "Sign in and grant data-plane access" section
    (Foundry project Responses requires AIServices/responses/*; Azure AI
    Developer is not sufficient; role table with Foundry Project Runtime User
    / Foundry User / Cognitive Services User; an az role assignment create
    example; and 401/403/404 troubleshooting bullets).

• What is the impact of these changes?

  • Samples run against a real Foundry project with just FOUNDRY_PROJECT_ENDPOINT
    (and optionally FOUNDRY_MODEL) in a .env — no source edits required.
  • Samples 05 and 06 no longer raise AttributeError: 'FunctionalWorkflow' object has no attribute 'build' on the published package.
  • No public API changes and no changes outside python/samples/01-get-started/.

• What do you want reviewers to focus on?

  • Whether gpt-4.1-mini is the right default model to advertise for the
    getting-started flow.
  • Whether the RBAC guidance in the README matches the current Foundry story
    (specifically the claim that Azure AI Developer alone is insufficient for
    project-scoped Responses and that AIServices/responses/* is required).
  • Sanity check of the FunctionalWorkflow .build() removal against the
    current shape of @workflow in python/packages/core/agent_framework/_workflows/_functional.py.

Related Issue

Fixes #

Contribution Checklist

• The code builds clean without any errors or warnings
• All unit tests pass, and I have added new tests where possible
• The PR follows the Contribution Guidelines
• This PR is linked to an issue and there is no other open PR for this issue (see Related Issue above).
• This is not a breaking change. If it is a breaking change, add the breaking change label (or add "[BREAKING]" to the title prefix, before or after any language prefix) — a workflow keeps the label and title prefix in sync automatically.

[Copilot]

Pull request overview

Updates Python getting-started samples to use environment-based Foundry configuration and expands setup guidance.

Changes:

• Loads Foundry endpoint/model settings from .env.
• Removes workflow .build() calls.
• Adds Foundry authentication, RBAC, and troubleshooting documentation.

Reviewed changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
README.md	Documents environment configuration and Foundry RBAC.
01_hello_agent.py	Loads Foundry settings from environment variables.
02_add_tools.py	Adds environment-based client configuration.
03_multi_turn.py	Adds environment-based client configuration.
04_memory.py	Adds environment-based client configuration.
05_functional_workflow_with_agents.py	Runs the decorated workflow directly.
06_functional_workflow_basics.py	Runs the decorated workflow directly.

---

💡 Add a code-review agent skill for context-aware, tailored reviews. Learn more in the docs.