https://github.com/rwth-acis/api-testing-bot
REST API test modeling (chat-) bot usable on GitHub & in the CAE. Uses OpenAI Codex for natural language test description understanding.
https://github.com/rwth-acis/api-testing-bot
bot codex devops-bot
Last synced: 9 months ago
JSON representation
REST API test modeling (chat-) bot usable on GitHub & in the CAE. Uses OpenAI Codex for natural language test description understanding.
- Host: GitHub
- URL: https://github.com/rwth-acis/api-testing-bot
- Owner: rwth-acis
- Created: 2022-07-29T10:47:50.000Z (almost 4 years ago)
- Default Branch: main
- Last Pushed: 2023-07-26T07:16:46.000Z (almost 3 years ago)
- Last Synced: 2025-06-05T12:08:22.011Z (about 1 year ago)
- Topics: bot, codex, devops-bot
- Language: Java
- Homepage:
- Size: 234 KB
- Stars: 1
- Watchers: 16
- Forks: 1
- Open Issues: 5
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
API Testing Bot
## ⚙️ Setup/Usage
First, set up the bot in the [Social-Bot-Framework](https://github.com/rwth-acis/Social-Bot-Framework).
Both the bot model and the NLU training data can be found in the `bot-model` directory.
The bot uses a backend service which is included in this repository.
The easiest way to use the service is to build (or pull) the Docker image and run the service as a container.
The bot can be used in different ways and on different platforms:
### 1. Usage with CAE & RocketChat
If the bot is a member of a RocketChat channel that is linked to a CAE project, test cases can be modeled via chat.
The bot can be triggered by sending a message such as "model a test".
As soon as the modeling is completed, the test case is forwarded to the CAE and proposed there in the Test Editor.
Dependencies:
- [las2peer-project-service](https://github.com/rwth-acis/las2peer-project-service)
- [CAE](https://github.com/rwth-acis/CAE)
### 2. Usage on GitHub
The bot may be used within GitHub repositories by connecting it to a GitHub app.
Therefore, the app id and a private key are required.
Please note that the private key needs to be converted from PKCS#1 to PKCS#8.
In the bot model, the `Authentication Token` of the GitHub issue or pull request messenger needs to be set to `[App Id]:[App Private Key]`.
Use the private key in PKCS#8 format, but remove the first and last line before copying it.
Also, set the webhook URL of the app to `.../apitestingbot/github/webhook/{gitHubAppId}`.
Then, the bot can be triggered within issues/pull requests by sending a message such as "model a test".
Once a test case has been modeled, the bot generates Java JUnit test code and comments it.
In pull requests, the bot may also propose spec-based test cases if it can access the developed service's OpenAPI documentation.
Read the [repository adjustment guide](repo_adjustment_guide.md) for more information.
Dependencies:
- [CAE-Code-Generation-Service](https://github.com/rwth-acis/CAE-Code-Generation-Service)
- [api-test-gen-service](https://github.com/rwth-acis/api-test-gen-service)
## 🐳 Docker Environment Variables
Depending on the use case different environment variables are required:
| Environment Variable | Description | Required? |
|--------------------------|----------------------------------------------------------------------------------------------|---------------------|
| `BOT_MANAGER_URL` | REST API URL of Bot Manager Service ending with `/SBFManager`. | Yes |
| `CODEX_API_TOKEN` | Token for accessing the OpenAI/Codex API. | Yes |
| `CAE_BACKEND_URL` | REST API URL of CAE Model Persistence Service ending with `/CAE`. | Only in 1. Use-case |
| `GITHUB_APP_ID` | Id of GitHub app that the bot should use. | Only in 2. Use-case |
| `GITHUB_APP_PRIVATE_KEY` | Private key of GitHub app that the bot should use (already needs to be converted to PKCS#8). | Only in 2. Use-case |