Source candidates from LinkedIn

You name the role and the company; your agent searches LinkedIn and pulls full profiles for the top matches. The recipe below runs in Claude, n8n, or your own code on Veezee's LinkedIn tools.

Recipe#

1. No signup: mint a free key (POST /v1/keys/mint, no body) and call every step below under a shared budget of 200 credits per IP per day. Hit the cap, or need limit: 30? Buy a key at /upgrade; checkout credits the same key.
2. linkedin_get_company: resolve the target company once, by URL, slug, or numeric id. Returns the company's numeric id and URN for the search filter. Cost: 4 credits.
3. linkedin_search_people: filter by title and current_company (the numeric id from step 2), limit: 30. Returns matches, each flagged is_anonymous if private. Cost: 12 credits.
4. linkedin_get_profile: for the top 10 non-anonymous results, fetch with sections: ["experience", "education"]. Cost: 4 credits per call, 40 credits for 10.

Skip any result with is_anonymous: true; linkedin_get_profile cannot dereference a private profile. Treat the search match itself as confirmation and move on.

Total cost#

56 credits for one company lookup, one search at limit 30, and 10 profile fetches.

Limits worth knowing#

• Free-key calls and trial keys cap search at 10 results per call; the limit: 30 step needs a paid key.
• A current_company name is matched by LinkedIn natively and works at any limit; a past_company name or company URL caps the limit, so pass the numeric id (from linkedin_get_company) for those at limit: 30.
• One search at a higher limit costs less than several at a low limit; a cursor page is priced fresh by its own limit, so paging does not save credits.
• With a paid key, call get_usage (free, exempt from the rate limit) before a large batch to check balance; a free key has no account balance to check, but each response reports its own cost.

Run it#

Full reference: /docs. A ready-made agent setup prompt: /agent-setup/prompt.md.

Installable skill: the candidate-sourcing pack teaches your agent this whole job. Install every pack with npx skills add veezeehq/veezee-skills (/docs/skills).