Roadmap for Future Development of Hello-Agents: 4 Planned Milestones Explained

The DataWhale Hello-Agents project outlines four concrete milestones: releasing video tutorials, stabilizing the HelloAgents SDK at v1.0.0, conducting a community-driven survey, and launching a "From Zero to One" training module for custom agent development.

The open-source hello-agents repository by DataWhale China provides a comprehensive learning framework for AI-native agent development. According to the official documentation in README.md, the project maintains a structured evolution plan under the "下一步规划" (Next Steps) section. This roadmap for future development of hello-agents bridges educational content with production-ready tooling, mapping directly to specific implementation layers within the codebase.

Four Strategic Milestones in the Hello-Agents Roadmap

The current roadmap defines four specific deliverables that advance the project from documentation to full-stack framework status.

1. Video Course Series

Plans include releasing hands-on tutorials that walk users through agent design, implementation, and deployment workflows. These videos will complement the existing text-based chapters located in docs/chapter*/ and the English documentation in README_EN.md.

2. HelloAgents Framework v1.0.0

This milestone targets a stable, lightweight, OpenAI-native SDK release. The framework provides utilities for memory management, context handling, and protocol management, building upon the custom agent architecture demonstrated in code/chapter7/my_main.py. The v1.0.0 release promises broader compatibility and enriched tooling while maintaining the existing lightweight philosophy.

3. Community-Driven Survey

A structured questionnaire will collect learner feedback to shape the next generation of learning modules. This data-driven approach ensures the roadmap aligns with actual user needs rather than theoretical use cases.

4. "From Zero to One" Training Project

A comprehensive tutorial series teaching users to train custom agents from scratch. This module covers data collection strategies, fine-tuning techniques, and evaluation metrics, moving beyond pre-built agents to custom model training.

How the Roadmap Aligns with the Technical Architecture

The planned development reinforces the repository's four-layer architecture. Each milestone targets specific implementation layers:

• Core Framework: The self-developed Python SDK wrapping OpenAI APIs, implemented in Co-creation-projects/YYHDBL-HelloCodeAgentCli/core/agent.py and Co-creation-projects/YYHDBL-HelloCodeAgentCli/core/llm.py. The v1.0.0 release hardens this layer with stable APIs and improved error handling.

• Framework Extensions: Built-in tools located in Co-creation-projects/YYHDBL-HelloCodeAgentCli/tools/builtin/*.py, including TODO, Terminal, Search, Memory, and MCP wrapper implementations. The roadmap promises additional utilities for rapid prototyping.

• Case Studies: Real-world implementations such as travel assistants and deep-research agents residing in code/chapter13/ and code/chapter14/. The v1.0.0 SDK maintains backward compatibility with these existing demos.

• Documentation & Learning: Structured Markdown chapters (docs/*.md) and community contributions (Extra-Chapter/*.md) expanding into multimodal video content per the first milestone.

Working with the HelloAgents SDK: Code Example

Developers can experiment with the current SDK implementation before the v1.0.0 release. The following example demonstrates initializing an agent with memory capabilities using the core framework classes:

from core.agent import Agent
from tools.builtin.memory_tool import MemoryTool
from core.llm import OpenAIChat

# Initialize the LLM client (requires OPENAI_API_KEY environment variable)

llm = OpenAIChat(model="gpt-4o-mini")

# Configure an agent with short-term memory storage

agent = Agent(
    name="demo-agent",
    llm=llm,
    tools=[MemoryTool()],
)

# Execute a query against the agent

response = agent.run("What are the three main layers of the Hello-Agents architecture?")
print(response)

This implementation references three key modules:

• Co-creation-projects/YYHDBL-HelloCodeAgentCli/core/agent.py – The Agent class orchestrating tool execution and LLM interaction.
• Co-creation-projects/YYHDBL-HelloCodeAgentCli/core/llm.py – The OpenAIChat wrapper handling API communication.
• Co-creation-projects/YYHDBL-HelloCodeAgentCli/tools/builtin/memory_tool.py – The MemoryTool class providing in-memory storage for context retention.

Summary

• The hello-agents roadmap defines four milestones: video tutorials, HelloAgents framework v1.0.0, community feedback survey, and the "From Zero to One" training project.
• The v1.0.0 release targets the Core Framework layer while maintaining compatibility with existing case studies in code/chapter13/ and code/chapter14/.
• Key implementation files include README.md (roadmap definition), Co-creation-projects/YYHDBL-HelloCodeAgentCli/core/agent.py (agent orchestration), and Co-creation-projects/YYHDBL-HelloCodeAgentCli/tools/builtin/*.py (tool implementations).
• Current capabilities can be tested using the provided Python SDK example with built-in memory and LLM tools.

Frequently Asked Questions

What is the expected release date for HelloAgents framework v1.0.0?

The repository documentation does not specify exact release dates for the v1.0.0 milestone. According to the README.md roadmap, this is categorized as a short-term priority focusing on SDK stabilization, enhanced tool utilities, and broader OpenAI API compatibility. Users should monitor the repository releases page for version updates.

How can I contribute to the hello-agents roadmap?

Contributors can participate through the planned community-driven survey mentioned in README.md to provide feedback on learning priorities. Additionally, the repository accepts code contributions via the Co-creation-projects/ directory (community extensions) and documentation improvements through docs/ and Extra-Chapter/ markdown files.

What specific skills will the "From Zero to One" training project teach?

According to the roadmap documentation, this training module covers the complete pipeline for custom agent development: data collection methodologies, model fine-tuning procedures, and evaluation frameworks. This moves beyond using pre-built agents to training proprietary models from scratch.

Is the HelloAgents SDK backward compatible with existing chapter implementations?

Yes. The roadmap explicitly states that the v1.0.0 release will solidify the Core Framework layer while preserving compatibility with existing case studies. The current demos in code/chapter13/ (travel assistant) and code/chapter14/ (deep-research agent) will continue functioning with the updated SDK architecture.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}