https://github.com/starwit/ai-cockpit-action-demo
A sample implementation to execute actions approved in AI Cockpit.
https://github.com/starwit/ai-cockpit-action-demo
Last synced: 5 months ago
JSON representation
A sample implementation to execute actions approved in AI Cockpit.
- Host: GitHub
- URL: https://github.com/starwit/ai-cockpit-action-demo
- Owner: starwit
- License: gpl-3.0
- Created: 2025-02-15T11:49:56.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-09-09T13:27:38.000Z (9 months ago)
- Last Synced: 2025-09-09T16:35:09.146Z (9 months ago)
- Language: Python
- Size: 272 KB
- Stars: 0
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- License: LICENSE
Awesome Lists containing this project
README
# AI Cockpit Action Executor
This component is part of project AI Cockpit, [see here](https://github.com/starwit/ai-cockpit-deployment) for a an overview of all components.
Actions selected and defined by AI componets are approved in AI Cockpit. Thus in order to execute them, decisions made in AI Cockpit needs to be communicated into the system landscape, that executes whatever AI has proposed.
Here is a sample implementation of all necessary functions, to forward AI Cockpit's decisions.
## What does it do
Service offers REST endpoints, to trigger actions and to feedback execution status.
## How to run locally
Service is written in Python and packaged in a Docker container. It is using Poetry package manager and to run application locally, use the following commands
```bash
poetry install
poetry shell
export SERVICE_URI=http://ip:8000 # if you want to access swagger ui from outside, default is localhost
python ai_cockpit_action_demo/main.py
```
You can reach API via: http://localhost:8000/docs (or http://ip:8000/docs)
Following Docker command runs executor on a Raspberry Pi (4 or newer, or you retire while waiting)
```bash
docker run -it --device /dev/gpiochip0 -e SERVICE_URI=http://ip:8000 -p 8000:8000 --rm starwitorg/ai-cockpit-action-demo:0.0.8
```
Without hardware connection, app will run with the following command.
```bash
docker run -it -e SERVICE_URI=http://ip:8000 -p 8000:8000 --rm starwitorg/ai-cockpit-action-demo:0.0.8
```
## How to install to Kubernetes
App can also be deployed to a Kubernetes cluster. Helm chart is located [here](https://hub.docker.com/r/starwitorg/ai-cockpit-action-demo-chart). See there for instructions how to deploy app.
## Contact & Contribution
The “KI-Cockpit” (AI Cockpit) project was funded by the Federal Ministry of Labor and Social Affairs.
# License
Software in this repository is licensed under the AGPL-3.0 license. See [license agreement](LICENSE) for more details.