wehub-resource-sync
e768098d0e
Flake8 Lint / flake8 (push) Waiting to run
Spell check CI / Spell_Check (push) Waiting to run
tools_continuous_delivery / Private PyPI non-main branch release (push) Has been skipped
tools_continuous_delivery / Private PyPI main branch release (push) Failing after 2m42s
Publish Promptflow Doc / Build (push) Has been cancelled
Publish Promptflow Doc / Deploy (push) Has been cancelled
1.5 KiB
1.5 KiB
Basic Chat — Microsoft Agent Framework
This is the Microsoft Agent Framework (MAF) version of the chat-basic Prompt Flow example.
It implements the same behaviour: a helpful assistant chatbot that remembers conversation history and responds to user questions.
Architecture
[InputExecutor] ──→ [ChatExecutor]
(question + (Agent with
chat_history) FoundryChatClient)
| Prompt Flow concept | MAF equivalent |
|---|---|
flow.dag.yaml |
WorkflowBuilder in chat_flow.py |
chat.jinja2 (system prompt) |
Agent(instructions="You are a helpful assistant.") |
LLM node (api: chat) |
FoundryChatClient + Agent.run() |
chat_history input |
Message list assembled in InputExecutor |
open_ai_connection |
Environment variables (FOUNDRY_PROJECT_ENDPOINT, FOUNDRY_MODEL) + DefaultAzureCredential |
Prerequisites
- Python 3.10+
- An Azure subscription with a Microsoft Foundry project (or Azure OpenAI resource)
az logincompleted
Setup
pip install -r requirements.txt
cp .env.example .env
# Edit .env with your Foundry project endpoint and model deployment name
Run
python chat_flow.py
This runs two test interactions:
- A single-turn question with no history
- A follow-up question with one prior turn of chat history