Able to switch off question recommendation

[wwwy3y3]

Is your feature request related to a problem? Please describe.
Question recommendation requires additional computation resources. Since this feature may not be needed in some scenarios, we could let users choose to disable.

Describe the solution you'd like
Bringing in a env variable that could turn off the question recommendation feature on following parts:

1. Home page: https://docs.getwren.ai/oss/guide/home/ask#recommended-questions
2. When questions require clarification: https://docs.getwren.ai/oss/guide/home/ask#out-of-scope-questions
3. When users question is a Database Schema Exploration: https://docs.getwren.ai/oss/guide/home/ask#database-schema-exploration
4. In follow-up questions: https://docs.getwren.ai/oss/guide/home/answer#recommended-questions

Additional context

1. also need an argument for the launcher to specify, ex: --diable-feature-recommendation

[flyrun9527]

support

[houjiaxxin]

support

[zc317414]

Strongly support ！

[Reza98Sh]

Hi

I’d like to work on this issue, but I’m unable to run the project successfully in my local setup.

Here’s my setup and the commands I’m using:

• Frontend:

  npm run dev
  

• AI service:

  poetry run uvicorn src.__main__:app --host 0.0.0.0 --port 5555 --loop uvloop --http httptools
  

• Other services:

  docker compose -f docker-compose-dev.yaml up qdrant wren-engine ibis-server
  

When I try to ask a sample question using the e-commerce test data, I get the following error:

POST /api/graphql 200 in 14ms
[2025-10-10T08:21:25.696] [ERROR] APOLLO - TypeError: Cannot read properties of null (reading 'hash')
at AskingService.getDeployId (webpack-internal:///(api)/./src/apollo/server/services/askingService.ts:634:27)
at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
at async AskingService.createAskingTask (webpack-internal:///(api)/./src/apollo/server/services/askingService.ts:293:26)
at async AskingResolver.createAskingTask (webpack-internal:///(api)/./src/apollo/server/resolvers/askingResolver.ts:174:22)
POST /api/graphql 200 in 16ms