Content

lemlist review and the Yalc Framework

The first outbound platform that ships its own Claude Code skill library plus an MCP. Yalc runs the whole loop from one prompt against a real lemlist account. Walkthrough page with the video and the orchestration skill at [/skills/lemlist-campaign-from-icp/](/skills/lemlist-campaign-from-icp/).

Yalc Fit Score
9/10
Pricing
from $59/mo
Trial
14 days free
Skills
38 open source
Last reviewed
2026-05-18
What it does

lemlist, plainly

lemlist is a multichannel outbound platform — email, LinkedIn, calls — with a built in people database, intent signals, and a 14 day free trial. In April 2026 it shipped two things that change how it pairs with Yalc: a library of 38 open source Claude Code skills (icp definition, persona definition, list building, sequence writing, reply handling, and more) plus an MCP server at `https://app.lemlist.com/mcp` that exposes most platform operations as Claude Code tools.

Stack alone, the 38 skills are atomic verbs. They take inputs, produce outputs, and forget. Yalc is the missing layer above them. Yalc owns the campaign state in SQLite, routes by seniority, picks the right lemlist skill for each lead, calls the lemlist MCP to push the campaign live, then loops replies back through lemlist's reply handler skill. From one Claude Code prompt, the whole outbound loop runs end to end against a real lemlist account.

The single vendor rule on this page: every node in the showcase workflow is lemlist. Source uses lemlist's People Database via the `lemleads_search` MCP operation. Enrichment is the new lemlist Agentic Enrichment that runs AI agents over every lead before any message is drafted. Writing uses lemlist's seniority routed copywriting skills. Sending is lemlist's own multichannel engine. Reply handling is lemlist's reply handler skill. Other lead sourcing options can plug in as swap in adapters, but they are not on the demo path.

Where it slots in

Position in the GTM operating system

Intake
Enrich
Score
Route
Draft
Send
Listen

lemlist sits across the **source**, **draft**, and **send** nodes of Yalc's topology, with its own loop back on **listen**. Yalc orchestrates the path between them, persists campaign state, and decides which lemlist skill fires for each lead. The lemlist MCP is the bridge that lets Yalc create campaigns, read stats, and pull reply events from inside Claude Code.

The Yalc Framework

Deploying lemlist inside a Yalc workflow

Workflow position

Multi node. Lead sourcing uses `people-finder` / `list-builder` / `company-finder` skills against lemlist's People Database. Reasoning uses `icp-definer` / `persona-definer` / `linkedin-outbound-angle`. Writing uses the seniority routed `copywriting-vp-sequence` / `manager-sequence` / `ic-sequence` plus `cta-designer` and `outbound-campaign-architect`. Sending hands off to lemlist's multichannel platform via the MCP. Replies route through `reply-handler`.

Prompt patterns

Copy paste prompts for Claude Code that invoke lemlist.

Yalc, run a lemlist intent signal trigger for the "HR tech VPs hiring AEs" segment. Use lemlist `people-finder` to source the list, run agentic enrichment, route by seniority through `persona-definer`, draft a three email VP sequence with `copywriting-vp-sequence` plus `cta-designer`, then push to lemlist as campaign "VP HRTech Q2." → Yalc handles routing and state, lemlist handles everything else.
Yalc, find 50 companies hiring sales engineers in Series B SaaS via lemlist `company-finder`. Filter to companies hiring in the last 14 days. Define personas with `persona-definer`, draft manager level sequences with `copywriting-manager-sequence`, push to lemlist as campaign "SE Hiring Manager." → No external data provider in this prompt. The whole graph runs against lemlist.
Yalc, on every new reply on the "VP HRTech Q2" lemlist campaign, run lemlist `reply-handler` to draft a response, surface the draft in Notion for human review, log the reply category for the weekly campaign report. → Yalc state plus lemlist primitives equals a reply loop you actually trust.

Chaining recommendations

Upstreamlemlist intent signals or `lemleads_search` (MCP) → lemlist `people-finder` / `list-builder` / `company-finder` skills → lemlist Agentic Enrichment
Downstreamlemlist reasoning skills (`icp-definer`, `persona-definer`, `linkedin-outbound-angle`) → lemlist copywriting skills (seniority routed) → lemlist MCP `create_campaign` → lemlist multichannel send → Yalc `track-campaigns` → lemlist `reply-handler`

Anti patterns to avoid

Don't ship a campaign from raw lemlist skills without state. The skills are stateless. Without Yalc tracking what's been sent to whom, you'll send touch one to the same VP three times. Pipe through Yalc so every send writes to SQLite.
Don't write copy with the seniority routing skills without first running `persona-definer`. The VP / Manager / IC split is the differentiator. Skipping persona reasoning means the seniority skills pick the wrong template, which costs reply rate.
Don't push to lemlist via MCP `create_campaign` without a dry run first. Yalc can render the full sequence locally and ask for a one prompt confirmation before the MCP call. Skipping that step has burned campaigns.
Don't disable lemlist Agentic Enrichment to save credits. The whole point of running this from Claude Code is that enrichment happens before the copywriting skills see the lead — letting them personalize on real signals. Disabling it makes the copy generic and the reply rate drops.

Yalc skill availability

Fourteen lemlist skills ship with Yalc out of the box under `.claude/skills/lemlist/`. They cover the full loop: `people-finder`, `list-builder`, `company-finder` (source), `icp-definer`, `persona-definer`, `linkedin-outbound-angle` (reason), `copywriting-first-touch`, `copywriting-follow-up`, `copywriting-vp-sequence`, `copywriting-manager-sequence`, `copywriting-ic-sequence`, `cta-designer`, `outbound-campaign-architect` (write), `reply-handler` (loop). The remaining 24 lemlist skills are available standalone from the upstream repo — install any of them with `npx github:l3mpire/claude-skills <skill-name> --project`. The lemlist MCP at `https://app.lemlist.com/mcp` is pre declared in Yalc's `.mcp.json` and reads `LEMLIST_API_KEY` from the environment. OAuth is the alternative for interactive use.

✓ Yalc skill available. View on GitHub.
Operator take

Pros, cons, who it's for

Pros

  • 38 open source Claude Code skills cover the full outbound loop, not just sequence writing
  • Native MCP server exposes campaign creation, lead search, stats, replies — same operations a paid user has in the UI
  • Agentic Enrichment shipped April 2026, AI agents research every lead before any message is drafted
  • Three published price tiers, 14 day free trial, no card to start
  • One vendor for source, enrich, write, send, reply. No 5 contracts to stitch together.
  • Skills are MIT, bundled into Yalc with full attribution. Update from upstream any time.

Cons

  • The whole point of running through Yalc is to keep state outside lemlist. Sending raw from lemlist's UI skips the routing logic, so the value of this integration only lands if you commit to Claude Code as the entry point.
  • The seniority routed copywriting skills require a real persona analysis first. Skip `persona-definer` and the copy reverts to generic.
  • Multichannel send (email plus LinkedIn plus calls) sits on the Multichannel Expert tier or higher. The Email Pro tier supports the skills and MCP but not the LinkedIn or call channels.
  • lemlist's intent signals are good but not as broad as a dedicated company signal provider (PredictLeads, Crustdata). For deep firmographic enrichment you'll still want to swap in an adapter.

Who it's for

  • GTM engineers who want one outbound vendor and one orchestration layer (Yalc), not five vendors and a Zapier graph
  • Founders running their own outbound from Claude Code who want a real platform underneath
  • Agencies managing multiple client campaigns who want the same Claude Code prompts to drive different lemlist accounts via `--tenant`
  • Operators evaluating whether AI native outbound is real — the 14 day free trial plus the 38 open source skills are the cheapest way to find out
Pricing reality

What you'll really spend

lemlist runs three published tiers. The Email Pro plan opens the door at $59 per user per month with email sending, warm up, and the core sequence engine. The Multichannel Expert plan steps up to $99 per user per month and adds LinkedIn outreach, calls, and the AI personalization features. The Outreach Scale plan at $159 per user per month adds advanced multi account, send reputation tools, and higher seat caps. The 38 Claude Code skills and the MCP server are available across all paid tiers, with the MCP exposing the same operations a paid user has in the UI. A 14 day free trial covers all tiers, no card required.

Email Pro

from $59

Email only. Sequence engine, warm up, send infrastructure. Right for solo operators or small teams running outbound from one mailbox.

Multichannel Expert

from $99

Adds LinkedIn, calls, AI personalization. The right tier for the Yalc workflow on this page since the demo uses multichannel sequences.

Outreach Scale

from $159

Multi account, advanced send reputation, higher seat caps. For agencies and teams running outbound at scale.

FAQ

Frequently asked

Do I need a paid lemlist plan to use the skills?

The 38 Claude Code skills are MIT licensed and free. The lemlist MCP server and the platform features it exposes (campaign creation, agentic enrichment, multichannel send) require a lemlist account. The 14 day free trial is enough to validate the full Yalc plus lemlist loop end to end before deciding on a tier.

What does the lemlist MCP server expose?

Most operations a paid user has in the lemlist UI — create / edit / pause / delete campaigns, retrieve campaign stats, search leads via `lemleads_search`, write sequences, import leads with deduplication, run a campaign audit. Ask the MCP directly inside Claude Code with "what lemlist operations can you perform?" for the current list.

Which 14 skills ship with Yalc out of the box?

Source — `people-finder`, `list-builder`, `company-finder`. Reason — `icp-definer`, `persona-definer`, `linkedin-outbound-angle`. Write — `copywriting-first-touch`, `copywriting-follow-up`, `copywriting-vp-sequence`, `copywriting-manager-sequence`, `copywriting-ic-sequence`, `cta-designer`, `outbound-campaign-architect`. Loop — `reply-handler`. The other 24 lemlist skills are out of GTM-OS scope (n8n, slide decks, cold call scripts, generic scrapers, CRM dedupe).

How does the MCP authenticate?

Two options. API key via the `X-API-Key` header, generated in lemlist Settings → Integrations and read by Yalc's `.mcp.json` from `LEMLIST_API_KEY` in the environment. Or OAuth via `claude mcp add --transport http lemlist https://app.lemlist.com/mcp` — tokens are valid for one hour and refresh automatically for 30 days.

Can I swap lemlist's lead sourcing for Crustdata or PredictLeads?

Yes. The single vendor rule on this page is for the showcase demo, not a hard architectural rule. Yalc adapters let you swap any source layer in. Use Crustdata or PredictLeads if you need broader firmographic or signal coverage than lemlist's People Database. The reasoning, writing, send, and reply layers stay on lemlist.

What is the Agentic Enrichment feature?

A new lemlist feature shipped in April 2026. AI agents automatically research every lead — website, LinkedIn, news, hiring activity — before any message is drafted. The output flows into the copywriting skills so personalization is grounded in real signals rather than firmographics alone.

Do the skills work without Yalc?

Yes. Install any skill directly with `npx github:l3mpire/claude-skills <name>` and use it standalone in Claude Code. Yalc adds orchestration, state, dedupe, and the campaign loop on top. The skills are useful alone, more useful together.

Run lemlist from Claude Code today.

Open source. Your data on your machine.