aws-samples / aws-samples/bedrock-chat
[Feature Request] Centralised internet search configuration
- Dominant language
- TypeScript
- Stars
- 1.3k
- Forks
- 535
- Avg merge
- 1d 12h
- Merged PRs (30d)
- 10
Description
## Describe the solution you'd like
Centralised way to configure the _internet search_ engine (DuckDuckGo or Firecrawl) and the credentials (Firecrawl API key) rather than having to configure it separately for each bot.
With this in place the Bot configuration can be simplified to Internet Search Yes / No and if yes how many documents to retrieve.
The search engine type and creds can be in Secrets Manager - we can create that with CDK and populate from cdk.json / parameters.ts defaults.
## Why the solution needed
Users should be able to create their own bots but don't necessarily need to have access to the required API keys and know which search engine to use.
## Implementation feasibility
Are you willing to collaborate with us to discuss the solution, decide on the approach, and assist with the implementation?
- [x] Yes, I am able to implement the feature and create a pull request.
Hope there's an appetite for this :)
Contributor guide
Research direction
Start by tracing the existing per-bot internet-search settings and the CDK/Secrets Manager configuration. Determine how search-engine selection and the Firecrawl API key are supplied from cdk.json or parameters.ts, then verify that bot configuration only needs the search toggle and document count while users do not need direct API-key access.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- aws
- Domain
- backend, cloud
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100