Gather leads from LinkedIn by searching 1st-degree connections matching specific criteria (role, seniority, location, industry). Use this skill whenever Shaw asks to find people on LinkedIn, pull connections matching a profile, build a lead list, gather contacts for outreach, find people for discovery calls, or anything involving searching through LinkedIn connections. Even casual mentions like "check my LinkedIn for analytics people" or "who do I know at mid-sized companies" should trigger this skill. Requires Chrome tools (Claude in Chrome) to be enabled.
This skill walks you through systematically extracting leads from a user's LinkedIn network. The core insight is that LinkedIn's connections page search only filters by name — to find connections by role or title, you need to use the main LinkedIn search bar with a 1st-degree filter.
If Chrome tools aren't available, tell the user to enable them at Settings → Desktop app → Computer use, and that they need to be logged into LinkedIn.
Before touching LinkedIn, nail down:
A single search term misses most of the network. You need multiple queries to get good coverage because people describe the same role in wildly different ways. For an analytics-focused search, good query sets look like:
Leadership tier:
Manager tier:
IC tier:
Location-specific (the geo hack):
Run the highest-seniority queries first. Stop adding queries once you've hit the target volume.
Don't try to keep 100+ leads in your context window. Create a JSON file to accumulate leads as you go:
/sessions/<session>/leads_raw.json
Write a small Python helper to add leads (deduplicating by URL) and report the running count. This pays for itself immediately — you'll be adding leads from 10+ pages across multiple searches.
For each search query: