brave

Overview ¶

Package brave wires Brave's Web Search API into web.Searcher.

Endpoint ¶

GET https://api.search.brave.com/res/v1/web/search

Authentication uses the X-Subscription-Token header (not bearer).

Parameter mapping ¶

web.SearchRequest → Brave query parameters:

• Query → q (after Google-style site:/-site: rewriting)
• MaxResults → count (default 10 when omitted)
• AllowedDomains → inlined as `site:foo.com` operators
• BlockedDomains → inlined as `-site:foo.com` operators
• Recency → freshness=pd/pw/pm/py (hour collapses to "pd" since Brave's minimum granularity is "past day")

Brave has no native allow/block-domain fields, so the query is rewritten with site:/-site: like the Serper provider does.

Response mapping ¶

Brave web.results[] → []*web.SearchResult:

• title → Title
• url → URL
• description → Snippet
• page_age → PublishedTime (RFC3339; the human-readable "age" field is dropped because relative strings like "2 hours ago" don't parse)

The provider deliberately owns and hides Brave's transport DTOs; only the normalized web result vertical crosses the package boundary.

Why Brave ¶

Brave runs its own independent index (not a Google/Bing reseller), so adding it gives the agent a result source that doesn't drift with the Google duopoly. Pricing is friendly: ~2000 queries/month free, sub-cent per query on paid tiers.

Reference ¶

https://api-dashboard.search.brave.com/app/documentation/web-search/get-started