Getting Help¶
Documentation¶
This documentation is your primary resource. Use the search bar (press / or s) to find specific topics.
GitHub Issues¶
For bugs, feature requests, or questions:
Before Opening an Issue¶
- Search existing issues - Your problem may already be reported
- Check the docs - The answer might be here
- Prepare a minimal example - Help us reproduce the issue
Bug Report Template¶
Markdown
## Description
[Clear description of the bug]
## Steps to Reproduce
1. Create backend with...
2. Call method...
3. Observe error...
## Expected Behavior
[What you expected to happen]
## Actual Behavior
[What actually happened]
## Environment
- pydantic-ai-backend version: X.X.X
- pydantic-ai version: X.X.X
- Python version: 3.XX
- OS: [e.g., macOS 14.0, Ubuntu 22.04]
- Docker version (if using DockerSandbox): X.X.X
Community Resources¶
Pydantic AI¶
pydantic-ai-backend is designed for use with Pydantic AI. Their documentation is an excellent resource:
Related Projects¶
- pydantic-deep - Full agent framework
- pydantic-ai-todo - Task planning toolset
- subagents-pydantic-ai - Multi-agent orchestration
- summarization-pydantic-ai - Context management
FAQ¶
Which backend should I use?¶
| Use Case | Backend |
|---|---|
| Unit tests | StateBackend (in-memory, fast) |
| Local CLI tools | LocalBackend (persistent files) |
| Multi-user web apps | DockerSandbox + SessionManager |
| Untrusted code | DockerSandbox (isolated) |
| Mixed sources | CompositeBackend (route by path) |
How do I run without Docker?¶
Use LocalBackend for local filesystem operations:
For testing, use StateBackend:
Python
from pydantic_ai_backends import StateBackend
backend = StateBackend() # In-memory, no side effects
How do I disable shell execution?¶
For LocalBackend:
For the console toolset:
How do I restrict file access?¶
Use the permission system:
Python
from pydantic_ai_backends import LocalBackend
from pydantic_ai_backends.permissions import READONLY_RULESET
backend = LocalBackend(root_dir="/workspace", permissions=READONLY_RULESET)
Or use allowed_directories:
Python
backend = LocalBackend(
root_dir="/workspace",
allowed_directories=["/workspace", "/shared"],
)
Docker container won't start¶
- Ensure Docker is running:
docker info - Check image exists:
docker images - Pull if needed:
docker pull python:3.12-slim - On Linux, check permissions:
sudo usermod -aG docker $USER
Contributing¶
We welcome contributions! See our Contributing Guide for details.