1. Connect the FullEnrich MCP to Claude

This skill needs only one MCP active in your Claude workspace:

• FullEnrich MCP at https://mcp.fullenrich.com/mcp

New to MCPs? Follow the FullEnrich MCP setup guide.

2. Install the skill in Claude

Click Download Skill above to grab the skill .zip file, then install it in Claude:

• Open Customize in Claude

• Click Create new skill

• Click the + button at the top

• Choose Create Skill

• Pick Upload a skill and select the .zip you just downloaded

3. Open the role with Claude

Activate the skill by telling Claude what you are hiring for. Examples:

• "I need to hire a Senior Backend Engineer, Python, in France"

• "Source me 15 Product Managers at scale-ups in London"

• "Find candidates for a Head of Sales role in Germany, B2B SaaS experience"

• "Aide-moi à sourcer un Senior Designer pour notre studio à Paris"

The skill picks it up from there, no need to fill any form.

4. Run the conversational intake

Instead of a fixed questionnaire, the skill conducts a short interview adapted to your brief. It opens with 3 to 4 questions and keeps probing wherever the answer is vague:

• "You said 'Senior Engineer'. Senior how, 5 years, 10 years? IC shipping features, leading a squad, or architecting systems?"

• "VP Engineering at a 15-person company is usually a strong tech lead who codes 60% of the time, is that what you need or a manager-of-managers?"

• "What is your company domain? I always exclude your own employees from the search." (mandatory question, never skipped)

• "Who had this role before, and why did they leave?"

The tone adapts: with an experienced recruiter the skill goes fast and technical, with a first-time hiring founder it is more pedagogical and explains why each question matters. If you say "just search", the skill pushes back politely: better intake = sharper shortlist.

5. Confirm the search and the credit cost

Once the brief is tight, the skill recaps what it understood (role + scope, must-have / nice-to-have skills, location, target company profile, exclusions, volume), validates the industry term against FullEnrich's official taxonomy, and runs a free preview search to confirm volume. Then it asks for explicit go on the cost:

"42 candidates match. Enriching the top 15 with email + phone will cost ~165 credits. You have 1,500. Proceed?"

      Email + phone are always enriched together for recruitment, the skill does not split them

6. Get the ranked shortlist with outreach plan

When the enrichment finishes, the skill ranks every candidate into 3 tiers and presents them in chat:

• 🟢 Tier 1, call first: candidates the skill personally recommends, with a mini profile per person (career path, education, key skills, why this fit specifically, recommended outreach channel)

• 🟡 Tier 2, worth considering: solid profiles with transferable value, what is missing or to validate

• 🟠 Tier 3, backup: partial matches, only contact if Tier 1 and 2 do not convert

For every Tier 1 candidate the skill recommends a channel based on the data found: email + phone = email first, follow up by phone in 48h. Email only = email then LinkedIn InMail after 3 days. Phone only = call directly. Neither = LinkedIn InMail with heavy personalization. A summary at the bottom shows total sourced, Tier 1 / 2 / 3 counts, email %, phone %.

Built-in safeguards: the skill always asks for your company domain and excludes your own employees from the search. It never enriches without an explicit go on the cost, never uses the API's default 10,000 limit, and always calibrates titles against company size (a "CTO" at a 10-person startup and a "CTO" at Google are different roles). It is also honest when the data is thin: if only 4 strong matches come out of a 15-candidate ask, you get the 4 plus a clear suggestion to broaden a filter, not 11 padded mediocre profiles.

What is next? After delivering the shortlist, the skill offers natural follow-ups: draft personalized recruitment outreach for the Tier 1 candidates, source more with adjusted filters, push the candidates to your ATS or CRM (HubSpot, Attio, Monday, Notion, Airtable), or export the full list as CSV.