Hugging Face's logo

Spaces:
wizofavalon
/
First_agent_template
Sleeping

App Files Community
First_agent_template
File size: 5,027 Bytes 
0128def
e3d7fcc
0128def
 
 
 
9f39c86
0128def
 
3ce50c4
e0cc0af
e8d4494
 
 
e3d7fcc
0128def
 
 
3334632
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
 
---
title: First Agent Template
emoji: ⚡
colorFrom: pink
colorTo: yellow
sdk: gradio
sdk_version: 5.23.1
app_file: app.py
pinned: false
tags:
- smolagents
- agent
- smolagent
- tool
- agent-course
---

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference


# 🤖 First Agent Project

Welcome to the **First Agent Project** — an elegant and modular AI agent system powered by [SmolAgents](https://github.com/huggingface/smol-agents), capable of solving complex tasks using structured reasoning, tool use, and code execution. Designed with usability and extensibility in mind, this project features a dynamic Gradio-based interface, plug-and-play tools, and a powerful HuggingFace model backend.

---

## ✨ Features

- 🔁 **Multi-Step Reasoning** with an agent that can plan, execute code, and invoke tools.
- 🌐 **Web Search Tool** via DuckDuckGo for live queries.
- 🧭 **Webpage Reader Tool** to fetch and parse web pages in Markdown.
- 📅 **Time and Weather Tools** for real-time timezone and temperature queries.
- 🖼️ **Image Generation** support using HuggingFace tool hub.
- 🎯 **Gradio UI Integration** with step-wise conversational tracing.
- 📦 Modular structure — easily extensible with custom tools.

---

## 🧠 Project Structure

```bash
.
├── app.py                  # Main entrypoint - defines tools, loads agent, and launches Gradio UI
├── Gradio_UI.py           # Elegant multi-step UI for interaction using Gradio
├── tools/final_answer.py        # Defines a final answer tool to conclude tasks
├── tools/visit_webpage.py       # Tool to visit a URL and extract readable markdown content
├── tools/web_search.py          # Tool for DuckDuckGo web search queries
├── requirements.txt       # Python dependencies
├── agent.json             # Agent configuration (tools, model, prompts, rules)
```

---

## 🧰 Built-in Tools

| Tool Name         | Description                                                                 |
|-------------------|-----------------------------------------------------------------------------|
| `web_search`      | Search the web using DuckDuckGo                                             |
| `visit_webpage`   | Visit and parse the content of a webpage as markdown                        |
| `get_temperature` | Get real-time temperature of a given city                                   |
| `get_current_time_in_timezone` | Get local time in a specified timezone                         |
| `final_answer`    | Returns the final response to a task                                        |
| `text-to-image`   | Generates images using HuggingFace image generation API (remote tool)       |

---

## 🚀 Quick Start

### 1. Install Dependencies

```bash
pip install -r requirements.txt
```

### 2. Run the App

```bash
python app.py
```

This will launch a Gradio interface where you can interact with your agent conversationally.

---

## 🌈 Sample Tasks

Try asking the agent:

- "What is the current temperature in Tokyo?"
- "Find the age of the Pope and raise it to the power of 0.36"
- "Search for latest news on Mars exploration"
- "Generate an image of a futuristic AI-powered city"
- "What’s the current time in New York?"

---

## 🛠️ Under the Hood

### 🧠 Agent Model
- Powered by [`Qwen/Qwen2.5-Coder-32B-Instruct`](https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct)
- HuggingFace API wrapper (`HfApiModel`) with streaming and code execution.

### 💬 Prompt Planning Logic
- Configured in `agent.json` with step-by-step templates, planning logic, and reasoning strategy.
- Structured prompt system encourages:
  - Thought → Code → Observation → Final Answer loops

---

## 🖼️ UI Highlights

- 💬 Chat-based interface with message tracking
- ⚙️ Shows tool usage logs, execution reasoning, and token statistics
- 📁 Supports file uploads (PDF, DOCX, TXT)
- 🧵 Maintains step context and memory

---

## 🔧 Customize or Extend

### Adding Your Own Tool

You can add custom tools easily by subclassing `Tool`. Example:
```python
from smolagents.tools import Tool

class MyTool(Tool):
    name = "my_tool"
    description = "Describe what your tool does here."
    inputs = {'input1': {'type': 'string', 'description': 'Your input'}}
    output_type = "string"

    def forward(self, input1: str) -> str:
        # Your logic
        return "Output result"
```

Then add it in `app.py` to the agent's tools list.

---

## 📜 License

This project is licensed under the Apache 2.0 License. See the [LICENSE](https://www.apache.org/licenses/LICENSE-2.0) file for details.

---

## 🌟 Acknowledgements

Built using:

- 🤖 [SmolAgents](https://github.com/huggingface/smol-agents)
- 🌐 [DuckDuckGo Search](https://pypi.org/project/duckduckgo-search/)
- 🧾 [Markdownify](https://pypi.org/project/markdownify/)
- 🔥 [Gradio](https://gradio.app/)
- 🧠 [Qwen Model on HuggingFace](https://huggingface.co/Qwen)