People
Find hiring candidates or prospects from a plain description, check the evidence for each match, and save the ones you want to the CRM.
Describe who you're looking for, like "founding engineers at seed-stage AI startups", and your agent finds them. Use it for hiring candidates, prospects, or anyone else your company wants to reach. Each person comes back with the evidence for or against your filters, so you judge fit yourself instead of trusting a score.
Where It Looks
- The people index, the default. It spends money.
- GitHub: top contributors to the repos you name. Free, but no work history.
- X: people who recently posted about the topic. It holds up to $1 per search and charges the actual cost.
Every search has a spending limit, $1 by default and at most $5. If the estimate is higher, the search stops before it runs.
Filters
Your agent can filter by words in someone's current job title (or words to
exclude), location, years of experience, and their employer's stage, like
seed or Series A. Checking company stage costs one company lookup per
employer, up to 25 per search. People past that come back unknown.
Nobody is dropped. Each person gets a pass, fail, or unknown for every
filter, with the value it compared. unknown means the data was missing.
Save People to the CRM
Someone already in your CRM comes back marked, so your agent updates that contact instead of adding a copy. Your agent saves new people as contacts, matched by their identities, like email or profile.
Commands
| MCP | CLI | API |
|---|---|---|
cofounder people search \
--query "backend engineers who write about streaming systems" \
--source x --min-years-experience 4 --max-cost-usd 2--source is exa (the default), github with --github-repo, or x.
Filters are --title-keyword, --exclude-title-keyword, --location,
--min-years-experience, --max-years-experience, and --company-stage.
To save someone new, write a person change with
cofounder crm records write, putting the kind and value of each of the
person's identities in its match.