mcpField
C
campaign-delivery

campaign-delivery

github.com/cold-iqMarketing & growth347 installsUpdated 2w ago
Repository

About campaign-delivery

campaign-delivery is a skill for planning and running outbound campaigns through a nine-phase process, from segment research and email drafting to deployment and performance review. It emphasizes writing the email before building lists, deduplicating before enrichment, and confirming before activation.

Key features

  1. 01Nine-phase campaign process
  2. 02Segment research and email drafting
  3. 03Targeting criteria and list building
  4. 04Lead scoring, deduplication, and enrichment
  5. 05Campaign deployment and analytics

Use cases

  1. 01Plan an outbound campaign from brief to launch
  2. 02Build and prioritize a prospect list
  3. 03Review campaign results against benchmarks

MCP support

It describes campaign work using ColdIQ endpoints, including Instantly campaign and lead operations; it does not state an MCP interface.

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/cold-iq/coldiq-marketplace-skills --skill campaign-delivery

After installing, type/campaign-deliveryin Claude Code to confirm it loaded.

Details

From the project README

Campaign Delivery

Build outbound campaigns with a 9-phase system (based on Jordan Crawford / Blueprint GTM). This is the orchestrator — each phase points to the skill and ColdIQ endpoint that does the work.

Golden rule: write the perfect email FIRST, then build the list that makes it true. Copy (Phases 2–3) comes BEFORE list building (Phase 5).

The 9 phases

Phase 0 — Kick-off (free). Define the tension triangle for the segment (see crawford-method). No spend yet.

Phase 1 — Context. Pull background on the ICP and a flagship customer. → POST /v1/limadata/research/ai-search · 0.3 cr · limadata.research.ai_search

Phases 2–3 — Copy (free). Write 5–10 angles, pick PQS/PVP, draft the perfect email + sequence (see crawford-method, cold-email-copy). QA against the 7-component + benchmark checklists.

Phase 4 — Targeting criteria (free). Translate the email into list criteria (company + contact + signal + exclusion filters). Validate filter values for free: → POST /v1/limadata/references/autocomplete · free · limadata.references.autocomplete

Phase 5 — List building (costs credits). Build the TAM from the criteria. → POST /v1/ai-ark/companies · per result · ai_ark.companies.search → POST /v1/apollo/organizations/search · 1 cr · apollo.organizations.search Layer in signals (see signal-detection): → GET /v1/signalbase/funding-signals · ? cr · signalbase.funding_signals (unverified)

Phase 5b — Score & tier (see tam-scoring), then dedup (see list-dedup) BEFORE paying to enrich.

Phase 6 — Contacts & enrichment. Pull contacts for T1/T2, find + verify emails (see contact-enrichment). → POST /v1/limadata/prospect/employees · 25 cr · limadata.prospect.employees → POST /v1/limadata/find/work-email · 1 cr · limadata.find.work_email → POST /v1/findymail/verify · 1 cr · findymail.verify

Phase 7 — Personalization & QA. Generate per-segment opening lines / ideas, QA the merge fields against real lead data.

Phase 8 — Deploy. Create the campaign with ColdIQ defaults, load leads, verify, then activate (see instantly-api — always confirm with the user before activating). → POST /v1/instantly/campaigns · free · instantly.campaigns.create → POST /v1/instantly/leads · free · instantly.leads.create → POST /v1/instantly/campaigns/{id}/activate · free · instantly.campaigns.activate

Phase 9 — Learn. Pull performance, compare to benchmarks, log what worked. → GET /v1/instantly/campaigns/analytics · free · instantly.campaigns.analytics

Timeline & principles

Copy + targeting (Phases 0–4) ≈ 2.5 hours of thinking before a single credit is spent. Spend nothing until the email is written. Dedup before enrichment. Test on a sample before the full send. Each follow-up must add new information (max 2–3 touches).

Per-campaign file structure

campaigns/{name}/  →  brief.md · copy/ · targeting.md · leads/ · report.md

Comments

Markdown supported