Note:
Code and instructions relating to the previous version of the tutorial Build a database-driven Slackbot are located in the branch cloud-functions.
Build a database-driven Slackbot with IBM Watson Assistant
This repository contains the code for IBM Cloud solution tutorial Build a database-driven Slackbot. It uses IBM Watson Assistant to create a chatbot and link it to Slack for a "Slackbot". By integrating a custom extension into the assistant, the chatbot can reach out to a Db2 on Cloud database. The custom extension is a Python app which provides a REST API for database objects. The chatbot can retrieve and insert data by calling these API functions. The Python app is deployed to IBM Code Engine.
Run and test locally
The app was tested with Python 3.10. You can either run the app directly or first build, then run it as container. Direct:
- Install the required modules as specified in requirements.txt.
- Edit .env to configure the database connection and a secret token / API key.
- Run
python3 app.py
Use the podman or docker tool to build and run the container.
docker build -t slackbot:2 .- Edit .env to configure the database connection and a secret token / API key.
docker run -p 8080:8080 --env-file .env slackbot:2
The file .env.sample shows a sample configuration:
- DB2_URI: Has the SQLAlchemy URL to the Db2 database
- API_TOKEN: Is the secret for the API key authentication.
- TABLE_ARGS: Is an optional setting to specify a different table schema. The EVENTS table is created in the default schema for the connection. This can be overwritten.
If you stored the Db2 credentials in a file slackbotkey.json, then use jq to create the URL using this command:
cat slackbotkey.json | jq '.[].credentials.connection.db2 | (.authentication.username + ":" + .authentication.password + "@" + .hosts[0].hostname + ":" + (.hosts[0].port | tostring) + "/" + .database + "?Security=SSL")'
The output follows this schema db2+ibm_db://user:[email protected]:port/bludb?Security=SSL;.
Create or recreate the database objects by calling the following API. Adapt the host, port, and API key as necessary.
curl -X 'POST' 'http://127.0.0.1:5000/database/recreate?confirmation=True' -H 'accept: application/json' -H 'API_TOKEN: MY_SECRET'
OpenAPI Specification
The Python app is based on the APIFlask API framework which itself is based on Flask. The OpenAPI specification for the application can be (re-)generated using:
flask spec --output slackbot-openapi-spec.json
You can also access the running app at the /docs (Swagger UI) and /redoc (Redoc) endpoints for interactive API documentation. See the APIFlask documentation on Swagger UI and Redoc for details.
License
See License.txt for license information.