> ## Documentation Index
> Fetch the complete documentation index at: https://docs.staffer.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Run a search

> Scout the corpus for candidates that match your job's criteria, in multiple rounds from the same canvas.

Open the **Scouting** tab on a job and describe who you're looking for, or let the agent search directly from the criteria already in the brief.

## How search works

Staffer searches its own candidate database using both the meaning of your criteria and keyword matches, then ranks results against the specific job. A candidate who's an excellent fit for one job and a poor fit for another gets two different scores, because the ranking is scoped to that job's criteria each time.

## Running rounds

Search happens in rounds, streamed into the canvas as they run. Ask for a different angle, for instance "look for people from smaller startups too" or "widen this to include contractors," and the agent runs another round without losing the first.

## Working the results from chat

You don't have to click through results one by one. Tell the agent what to do with them directly: "add the top 5," "remove anyone outside CET," "shortlist everyone above an 80." The canvas updates to match.

Results carry a status badge so you always know where a candidate already stands: shortlisted, already in your candidates list, or attached to a different job.

**Next:** [Reading match scores](/companies/match-scores)