gabyorel commited on
Commit
a6afbde
Β·
1 Parent(s): 4b38877

Update README

Browse files
Files changed (1) hide show
  1. README.md +72 -4
README.md CHANGED
@@ -1,14 +1,82 @@
1
  ---
2
  title: SafeBox
3
- emoji: πŸ“Š
4
- colorFrom: green
5
- colorTo: yellow
6
  sdk: gradio
7
  sdk_version: 5.33.0
8
  app_file: app.py
 
 
 
 
 
 
9
  pinned: false
10
  license: apache-2.0
11
  short_description: Execute Python code safely in a dynamic sandbox.
12
  ---
13
 
14
- Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  title: SafeBox
3
+ emoji: πŸ› οΈ
4
+ colorFrom: gray
5
+ colorTo: pink
6
  sdk: gradio
7
  sdk_version: 5.33.0
8
  app_file: app.py
9
+ tags:
10
+ - mcp-server-track
11
+ - code-execution
12
+ - modal
13
+ - sandbox
14
+ - dynamic-dependencies
15
  pinned: false
16
  license: apache-2.0
17
  short_description: Execute Python code safely in a dynamic sandbox.
18
  ---
19
 
20
+ # SafeBox MCP Server
21
+
22
+ This project is a Gradio-based MCP server that allows both users and AI agents to **safely execute arbitrary Python code snippets** in a Modal sandbox. It automatically **detects and installs Python dependencies** before executing the code.
23
+
24
+
25
+ ## ✨ Features
26
+
27
+ - **Sandboxed Code Execution**: All code is executed inside a secure, isolated Modal sandbox.
28
+ - **Dynamic Dependency Resolution**: Automatically detects imported packages in the code and installs them on the fly using `pip`.
29
+ - **Runtime Feedback**: Returns `stdout`, `stderr`, `returncode`, and any error messages from the execution.
30
+ - **Build Logs**: Returns Modal build logs for full visibility.
31
+
32
+
33
+ ## πŸ€– Designed for AI Agents
34
+
35
+ SafeBox was built with **autonomous AI agents** in mind.
36
+
37
+ It is ideal for agents that need to **reflect on code changes** by executing, analyzing, and iterating on their own generated Python code.
38
+
39
+
40
+ ## πŸ”Œ Integration with MCP clients
41
+
42
+ All you need to do is add this URL endpoint to your MCP Client (e.g., Cursor, Cline, or Tiny Agents), which typically means pasting this config in the settings:
43
+
44
+ ```json
45
+ {
46
+ "mcpServers": {
47
+ "safebox": {
48
+ "url": "https://agents-mcp-hackathon-safebox.hf.space/gradio_api/mcp/sse"
49
+ }
50
+ }
51
+ }
52
+ ```
53
+
54
+ Some MCP Clients, notably Claude Desktop, do not yet support SSE-based MCP Servers.
55
+ In those cases, you can use a tool such as mcp-remote. First install Node.js. Then, add the following to your own MCP Client config:
56
+
57
+ ```json
58
+ {
59
+ "mcpServers": {
60
+ "safebox": {
61
+ "command": "npx",
62
+ "args": [
63
+ "mcp-remote",
64
+ "https://agents-mcp-hackathon-safebox.hf.space/gradio_api/mcp/sse",
65
+ "--transport",
66
+ "sse-only"
67
+ ]
68
+ }
69
+ }
70
+ }
71
+ ```
72
+
73
+ ## πŸ” Authentication & Tokens
74
+ βœ… You do not need to set up your own Modal account to use this Space.
75
+
76
+ This Space uses a preconfigured Modal token, and you are free to use my credits to run your code snippets in a secure sandbox.
77
+
78
+ ## πŸ“Ί Demo Video
79
+ Curious to see it in action?
80
+
81
+
82
+ Watch this short demo showing how the sandbox runs Python code with automatic dependency detection and live output: