Skip to content

Scouting Sources

ThreadLens can scout social posts and Google Search results from project-specific queries. Useful scoring and reports require AI provider readiness even when a source itself has little or no extra credential setup.

SourceBest forSource credentialsFirst-run guidance
RedditLong-form complaints, niche community language, and detailed workaroundsNo extra source credential is documented for the current first-run pathStart here when you want the least source-specific setup. Configure an AI provider first for useful scoring and reports.
Google SearchPages, comparisons, search intent, commercial intent, and recurring problem language outside social postsPARALLEL_API_KEYUse after the configured search provider key is present in the repository root .env.
BlueskyShorter public posts, emerging conversation, and social commentaryBLUESKY_HANDLE and BLUESKY_APP_PASSWORDUse after both Bluesky credentials are present in the repository root .env.

Reddit is useful for long-form complaints, niche community language, and detailed workarounds. It is the recommended first source when you want the least source-specific credential overhead.

Reddit results still need AI provider readiness before you should expect useful scoring, analysis, or reports.

Bluesky is useful for shorter public posts and emerging conversation. Configure BLUESKY_HANDLE and BLUESKY_APP_PASSWORD before relying on Bluesky scouting.

Google scouting finds pages and search results that can show comparison intent, commercial intent, and recurring problem language. Configure PARALLEL_API_KEY when using the Parallel.ai Search provider.

Start with one source, inspect the results, then add more sources after the project query language is producing useful findings. If a source returns errors or empty results, check both the AI provider credential and the source-specific credential for that source.

For first-run credential setup, see Configuration Basics. For the complete variable reference, see Environment Variables.