agentleFS
automatewithudayautomatewithuday/gtm-prospecting-system/skills/contact-enrichment/SKILL.md

contact-enrichment

Skill1 starsChanged 59 days agoMITView on GitHub

Use when stage 8 of the prospecting system needs people and verified emails — all matching titles at the approved accounts, plus work emails for qualified audience members who only have a LinkedIn URL — merged into one deduped lead file. Also use when the ask is \"find the contacts and emails for these companies\" or \"get emails for these LinkedIn profiles\".

  • Reads credentials

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    No LinkedIn URL but name + company → getleads_enrich_person_batch.

    line 41

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Contact Enrichment

Two lanes arrive here with different shapes: accounts that need people, and people that need emails. One file leaves.

Pipeline contract

Stage 8 of 20. Reads 07-accounts.csv + 07-lane.json (account lane) and 06-audience-qualified.csv (audience lane). Writes 08-leads.csv with the lead columns.

Files live in runs/<run>/. Column definitions: /prospecting-system.

Gate: the human approves 07-accounts.csv before any contact pull. This is where real money starts.

Lane A — accounts → people

npx tsx engine/list-builder/scripts/contacts.ts --config=<run>/07-lane.json --run-dir=<abs path to run>/07-lane

Cost-ordered waterfall, resumable: GetLeads export (free on plan) → Blitz for domains GetLeads left at zero → Prospeo people search last (paid). It refuses to run unless the lane's summary.md says READY. Output is leads-final.csv in the lane dir.

The engine pulls all matching titles at every company, uncapped — that is its rule, keep it. The run's target is enforced later, at stage 11, by ranking; not here by dropping contacts.

Set source_type to lookalike for accounts that came from 03-lookalikes.csv, else search.

Lane B — audience members → emails

Rows in 06-audience-qualified.csv with qualified=true and no email:

  1. Has linkedin_url → GetLeads MCP getleads_get_emails_from_linkedin_batch.
  2. No LinkedIn URL but name + company → getleads_enrich_person_batch.

Both charge one plan credit per successful row. Check get_fair_use first on large batches. Only enrich qualified rows — never the raw audience.

Merge

  1. Normalize both lanes to the lead columns. Build lead_id.
  2. Dedupe by linkedin_url, else account_domain+first_name+last_name. When the same person arrives from both lanes, keep ONE row, keep the audience source_type (it is the stronger evidence), and add the sources up in source_count.
  3. Drop excluded_domains and anyone on the do-not-contact list.
  4. Clean names before anything is written to a person. Provider names arrive as Newtekone Nasdaq Newt, Performyard, Extensishr. Add company_name_clean and first_name_clean using references/company-name-cleaning.md and references/first-name-cleaning.md. Stages 13–16 merge the clean columns, never the raw ones.

Email verification — the send gate

A provider saying "verified" is a hint. Every address goes through MillionVerifier before it is send-ready (script and key setup: cold-email-starter-kit).

  • email_status=valid only for MillionVerifier ok.
  • Catch-all, unknown, and invalid are kept in the file with their status and are never emailed. They can still get LinkedIn outreach in stage 15.
  • MillionVerifier rejects default Python/Node user agents — send a browser UA.

Verification is paid per address: add it to spent_usd first.

Report

Rows per lane, duplicates merged, emails found, emails valid, cost. If valid emails are under the run's target, say so now — the fix is more accounts or another audience, not looser verification.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.