Keyword search puts the page that repeats your words on top, not the page that actually answers you. This layer sits over Brave Search and re-orders results by meaning.
sonnet4 daysBrave Search APIGeminipgvectorFastAPI
Claude
67ROI
75Scale
$1.9k93Saved
ROI for
README.md
Why this subagent
Keyword search puts the page that repeats your words on top, not the page that actually answers you. This layer sits over Brave Search and re-orders results by meaning.
It embeds the query and every result, then scores them for real intent match before handing back a cleaner ranking. Drop it behind any agent or app that needs better search than raw keyword results, with a relevance note attached to each hit.
How it runs
Used at step 01 to kick off the pipeline.
Write
Used at step 01 to kick off the pipeline.
WebFetch
Used at step 01 to kick off the pipeline.
WebSearch
Used at step 01 to kick off the pipeline.
Embed the query and each result, then score them for true semantic relevance.
pending
Re-order the results so the most on-intent pages rise above keyword-matched ones.
pending
Return the re-ranked list with a short relevance note per result for any agent to use.
pending
Sample output
json
// Sample output
// (generated when the pipeline finishes)
Re-rank the Brave results by semantic match to the query intent and return them ordered, each with a relevance_score and one-line justification.
Unlock the rest
The full agent definition, install snippet, and starter task are gated for community members.
Members get the full `.md` agent file, the npm / pnpm install one-liners, a starter prompt that we've tuned against real runs, and the open-source repo when this automation ships there. One email, magic link, done.