Migrate deepseek-qwen-autogen-agent from Magic to Pixi (#68)

* Migrate deepseek-qwen-autogen-agent from Magic to Pixi

## Summary
- Replaced Magic CLI with Pixi package manager throughout the recipe
- Replaced global max-pipelines installation with modular package dependency
- Updated all references from "MAX Serve" to "MAX" for consistency

## Changes
- .gitignore: Updated from "magic environments" to "Pixi environments", added pixi.lock
- README.md:
  - Replaced Magic installation instructions with Pixi
  - Removed `magic global install -u max-pipelines` command
  - Changed all `magic run` commands to `pixi run`
  - Updated from `magic init` to `git clone` for recipe download
  - Renamed "MAX Serve" to "MAX" throughout
- pyproject.toml:
  - Added `modular = ">=25.5.0.dev2025070905,<26"` dependency to replace max-pipelines
  - Reordered channels to prioritize max-nightly channel
- metadata.yaml:
  - Updated tasks from `magic run` to `pixi run`
  - Updated titles from "MAX Serve" to "MAX"

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

* Fix system requirements link to use FAQ page

Updated system requirements reference to point to the FAQ page as per PR 65 format:
https://docs.modular.com/max/faq/#system-requirements

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

* Fix rich dependency version conflict in deepseek-qwen-autogen-agent

Update rich dependency constraint from <14 to <15 to resolve CI test failures caused by version conflict with rich==14.0.0.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

---------

Co-authored-by: Claude <noreply@anthropic.com>

Author: williamw

deepseek-qwen-autogen-agent/.gitignore

@@ -1,6 +1,5 @@
 
-# pixi environments
+# Pixi environments
 .pixi
 *.egg-info
-# magic environments
-.magic
+pixi.lock

deepseek-qwen-autogen-agent/README.md

@@ -1,10 +1,10 @@
-# Learn How to Build AI Agents with DeepSeek-R1, AutoGen and MAX Serve
+# Learn How to Build AI Agents with DeepSeek-R1, AutoGen and MAX
 
 This recipe demonstrates how to build AI agents using:
 
 * [`DeepSeek-R1-Distill-Qwen-7B`](https://builds.modular.com/models/DeepSeek-R1-Distill-Qwen/7B) model that runs on GPU
 * [AutoGen](https://microsoft.github.io/autogen/stable/) framework for multi-agent conversations
-* [MAX Serve](https://docs.modular.com/max/serve/) for efficient model serving and inference
+* [MAX](https://docs.modular.com/max/serve/) for efficient model serving and inference
 * [Rich](https://rich.readthedocs.io/en/stable/introduction.html) Python library for beautiful terminal interfaces
 
 We'll create two example applications that showcase:
@@ -22,24 +22,18 @@ The patterns demonstrated here can be adapted for various agent-based applicatio
 
 ## Requirements
 
-Please make sure your system meets our [system requirements](https://docs.modular.com/max/get-started).
+Please make sure your system meets our [system requirements](https://docs.modular.com/max/faq/#system-requirements).
 
-To proceed, ensure you have the `magic` CLI installed with the `magic --version` to be **0.7.2** or newer:
+To proceed, ensure you have the `pixi` CLI installed:
 
 ```bash
-curl -ssL https://magic.modular.com/ | bash
+curl -fsSL https://pixi.sh/install.sh | sh
 ```
 
-or update it via:
+...and updated to the latest version:
 
 ```bash
-magic self-update
-```
-
-Then install `max-pipelines` via:
-
-```bash
-magic global install -u max-pipelines
+pixi self-update
 ```
 
 ### GPU requirements
@@ -52,33 +46,33 @@ This recipe requires a GPU with CUDA 12.5 support. Recommended GPUs:
 
 ## Quick start
 
-1. Download the code for this recipe using the `magic` CLI:
+1. Download the code for this recipe:
 
     ```bash
-    magic init deepseek-qwen-autogen-agent --from modular/max-recipes/deepseek-qwen-autogen-agent
-    cd deepseek-qwen-autogen-agent
+    git clone https://github.com/modularml/max-recipes.git
+    cd max-recipes/deepseek-qwen-autogen-agent
     ```
 
-2. Run the MAX Serve server via in a terminal:
+2. Run the MAX server via in a terminal:
 
     **Make sure the port `8010` is available. You can adjust the port settings in [pyproject.toml](./pyproject.toml).**
 
     ```bash
-    magic run server
+    pixi run server
     ```
 
 3. In a new terminal, run either example:
 
     * For the chat agent:
 
         ```bash
-        magic run chat_agent
+        pixi run chat_agent
         ```
 
     * For the screenplay development team:
 
         ```bash
-        magic run screenplay_agents
+        pixi run screenplay_agents
         ```
 
 The agents will be ready when you see the welcome message in your terminal.
@@ -138,7 +132,7 @@ assistant = AssistantAgent(
 
 Key features:
 
-* Uses DeepSeek-R1 model through MAX Serve
+* Uses DeepSeek-R1 model through MAX
 * Configurable temperature for response creativity
 * Adjust `max_tokens` for longer conversations
 
@@ -399,7 +393,7 @@ This recipe can be adapted for various applications:
 Common issues and solutions:
 
 1. **Server Connection Issues**
-   * Ensure MAX Serve is running (`magic run server`)
+   * Ensure MAX is running (`pixi run server`)
    * Check if the default port `8010` is available
    * Verify network connectivity
 

deepseek-qwen-autogen-agent/metadata.yaml

@@ -1,6 +1,6 @@
 version: 1.0
-long_title: "Learn How to Build AI Agents with DeepSeek-R1, AutoGen and MAX Serve"
-short_title: "AI Agent with DeepSeek-R1, AutoGen and MAX Serve"
+long_title: "Learn How to Build AI Agents with DeepSeek-R1, AutoGen and MAX"
+short_title: "AI Agent with DeepSeek-R1, AutoGen and MAX"
 author: "Ehsan M. Kermani"
 author_image: "author/ehsan.jpg"
 author_url: "https://www.linkedin.com/in/ehsanmkermani/"
@@ -15,6 +15,6 @@ tags:
   - autogen
 
 tasks:
-  - magic run server
-  - magic run chat_agent
-  - magic run screenplay_agents
+  - pixi run server
+  - pixi run chat_agent
+  - pixi run screenplay_agents

deepseek-qwen-autogen-agent/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 authors = [{ name = "Modular Inc", email = "hello@modular.com" }]
-dependencies = ["autogen-agentchat==0.4.7", "autogen-ext[openai]==0.4.7", "rich>=13.9.4,<14", "python-chess>=1.999,<2", "tenacity>=9.0.0,<10", "requests>=2.32.3,<3"]
+dependencies = ["autogen-agentchat==0.4.7", "autogen-ext[openai]==0.4.7", "rich>=13.9.4,<15", "python-chess>=1.999,<2", "tenacity>=9.0.0,<10", "requests>=2.32.3,<3"]
 name = "deepseek-qwen-autogen-agent"
 requires-python = ">=3.10,<3.13"
 version = "0.0.0"
@@ -16,12 +16,15 @@ requires = ["hatchling"]
 packages = ["."]
 
 [tool.pixi.project]
-channels = ["conda-forge", "https://conda.modular.com/max-nightly"]
+channels = ["https://conda.modular.com/max-nightly", "conda-forge"]
 platforms = ["linux-64", "linux-aarch64", "osx-arm64"]
 
 [tool.pixi.pypi-dependencies]
 deepseek_qwen_autogen_agent = { path = ".", editable = true }
 
+[tool.pixi.dependencies]
+modular = ">=25.5.0.dev2025070905,<26"
+
 [tool.pixi.tasks]
 server = "MAX_SERVE_PORT=8010 max serve --model-path deepseek-ai/DeepSeek-R1-Distill-Qwen-7B --max-length 16384 --max-batch-size 1"
 chat_agent = "python chat_agent.py"