Python scripts to simulate enough of the openwebui environment to run openwebui pipe functions directly.
|
|
||
|---|---|---|
| .forgejo/workflows | ||
| tests | ||
| .gitignore | ||
| .pre-commit-config.yaml | ||
| AGENTS.md | ||
| openwebui_memory_access.py | ||
| openwebui_pipe_interface.py | ||
| openwebui_standard_functionality.py | ||
| pipe.py | ||
| README.md | ||
| runner.py | ||
| webui_runner.md | ||
OpenWebUI Runner
Overview
The OpenWebUI runner provides a command-line interface to execute OpenWebUI-compatible pipe functions outside of the full OpenWebUI environment.
Purpose
This runner allows testing and execution of OpenWebUI pipes with a simplified interface that mimics the actual OpenWebUI environment.
Components
- Runner Script: Executes pipe functions with mocked OpenWebUI environment
- Environment Mocks: Simulates OpenWebUI imports and services
- CLI Interface: Accepts pipe files and parameters via command line
Usage
python runner.py <pipe_file.py> [arguments]
Environment Simulation
The runner provides mocked versions of:
open_webui.models.usersopen_webui.utils.chat- FastAPI Request objects
- User context data
Limitations
- Does not connect to actual LLM services
- No real file/knowledge base access
- Limited to basic pipe functionality
- Mocked responses only
Requirements for Pipes
Pipes must implement the standard OpenWebUI pipe interface:
pipe()method withbody,__user__,__request__parameters- Optional
Valvesconfiguration class - Async support for streaming responses