feat(api): integrate external apps with nesis

[mawandm]

We introduce integrating external apps with Nesis by providing the apps API, which allows for adding an app and authenticating with an app secret.

Part of #83

[mawandm]

Tested with

1. Create an Admin user session with

   NESIS_TOKEN=$(curl -s -X POST -H "Content-Type: application/json"  -k http://localhost:6000/v1/sessions -d 
   '{"email":"test.user@domain.com","password":"password"}' | jq -r '.token')
   

2. Create a datasource with

   curl -d '{"type": "s3", "name": "ht-documents", "connection": {"user": "your_username","password":"your_password","endpoint": "http://localhost:4566","dataobjects": "documents","region": "us-east-1"}}' -H "Content-Type: application/json"  -H "Authorization: Bearer $NESIS_TOKEN" -X POST http://127.0.0.1:6000/v1/datasources`
   

3. Post a document to the RAG engine with

   curl -X POST -H "Content-Type: multipart/form-data; boundary=----------------------------4ebf00fbcf09" --form "file=@Downloads/reseller-agreement.docx" --form 'metadata={"file_name":"Downloads/reseller-agreement.docx","datasource": "ht-documents"}' http://localhost:8080/v1/ingest/files
   

4. Create a role with

   curl -X POST --url http://localhost:6000/v1/roles -H "Content-Type: application/json" -H "Authorization: Bearer $NESIS_TOKEN" --data '{"name":"document-manager","policy":{"items":[{"action":"read","resource":"datasources/*"},{"action":"create","resource":"predictions/*"}]}}'
   

5. Attach the role to another user (previously created) with

   curl -X POST --url http://localhost:6000/v1/users/<user-id>/roles -H "Content-Type: application/json" -H "Authorization: Bearer $NESIS_TOKEN" --data '{"id":"<role-id>"}'`
   

6. Create an app with

   NESIS_APP_TOKEN=$(curl -X POST --url http://localhost:6000/v1/apps -H "Content-Type: application/json" -H "Authorization: Bearer $NESIS_TOKEN" --data '{"name":"test-app"}' | jq -r '.token')
   

7. Create a prediction as the app with (without the X-Nesis-Request-UserKey fails due to permissions)

   curl -X POST http://localhost:6000/v1/modules/qanda/predictions --header "Content-Type: application/json"  -H "Authorization: Bearer $NESIS_APP_TOKEN" --data '{"query": "Summarise only office agreement"}'
   

8. Create the prediction as the user with

   curl -X POST http://localhost:6000/v1/modules/qanda/predictions --header "Content-Type: application/json"  -H "Authorization: Bearer $NESIS_APP_TOKEN" -H "X-Nesis-Request-UserKey: <user-id>" --data '{"query": "Summarise only office agreement"}'