The 25 Best Claude Skills for GTM in 2026 (Tested & Compared)

1. positioning-icp (agent-gtm-skills), best for locking positioning and ICP

Overview. A dense positioning playbook that forces an agent to produce an ICP definition, a competitive position and a messaging architecture before it writes a single outbound line.

Key features. ICP definition with enrichment signals, competitive positioning frameworks, messaging architecture, and a treatment of perishable product-market fit.

Pros and cons. Unusually specific frameworks for a free skill file, and it runs anywhere with no dependencies. Against that, the repository has two commits, both from 13 February 2026, and carries no licence file, so a company that vets dependencies will struggle to approve it.

Pricing. Free. No licence file in the repository despite the MIT claim on the project site.

Best for. Founders and heads of marketing who need a defensible position written down before outbound starts.

2. Won-Deal ICP Finder (La Growth Machine), best for finding your proven ICP

Overview. Audits closed-won deals to surface the ICP you actually sell to rather than the one on the strategy slide, identifies the channel that won each archetype, and outputs lookalike searches per segment.

Key features. Closed-won audit across the CRM, archetype clustering, winning channel per archetype, and lookalike searches ready to run.

Pros and cons. It answers the question with evidence instead of opinion, and the output feeds straight into list building. It needs a connected account and enough closed-won history to be meaningful, so a team in its first quarter of outbound will get thin results.

Pricing. Free and MIT licensed. Running it needs a La Growth Machine account, from €60/month per identity on Basic with monthly billing, verified August 2026.

Best for. Teams with two or more quarters of closed-won data who suspect their stated ICP and their real one have drifted apart.

→ Won-Deal ICP Finder

3. Sales Nav Search Builder (La Growth Machine), best for turning an ICP into a search

Overview. Converts a plain-language target description into a ready-to-click Sales Navigator search, with boolean title strings, clean filters and a one-click import into an audience.

Key features. Boolean title construction, filter selection that avoids the usual Sales Navigator traps, geography and headcount handling, and direct import.

Pros and cons. It removes the part of list building people get wrong most often, which is boolean syntax. It assumes a Sales Navigator seat, and the import step needs a connected account.

Pricing. Free and MIT licensed. Requires a Sales Navigator seat, plus a La Growth Machine account for the import.

Best for. SDRs and GTM engineers who build lists weekly and lose time to boolean syntax.

→ Sales Nav Search Builder

4. prospect-discovery-pipeline (YALC GTM-OS), best for end-to-end account sourcing

Overview. Runs a full sourcing pass, from a described segment through to a scored account list, chaining company discovery, research and qualification inside one skill.

Key features. Company discovery, per-account research, an explicit qualification pass, and handoff into the rest of the GTM-OS routines such as find-lookalikes and qualify-leads.

Pros and cons. The chaining is well designed and the qualification gates are explicit. It is CLI-first rather than a drop-in Claude.ai skill, it needs provider API keys to return anything, and it belongs to the 1.0 generation its authors have declared feature complete.

Pricing. Free, MIT licensed. Data provider keys are billed separately by those providers.

Best for. GTM engineers comfortable in a terminal who want a sourcing pipeline they can fork and self-host. Pair it with dedicated AI prospecting tools when the data layer underneath needs to be stronger.

5. signal-sourcer (ColdIQ), best for intent and trigger signals

Overview. A master skill covering buying signals end to end: which triggers are worth watching, how to source them, and how to turn a trigger into an outbound angle rather than a generic mention of a funding round.

Key features. Trigger taxonomies, inbound and outbound trigger sets, signal-to-angle mapping, and sub-skills covering funding, hiring, job changes, tech changes and website visitors.

Pros and cons. The depth on signals is the best in this list by a wide margin. The repository ships no licence file, and its COLDIQ-FIRST.md mandate points the default execution path at ColdIQ’s own paid MCP server, with other tools as labelled fallbacks.

Pricing. Free to read. The documented execution path needs a paid ColdIQ API key.

Best for. Outbound teams whose reply rates have flattened on static lists and who need a trigger layer.

6. Multichannel Campaign Builder (La Growth Machine), best for writing the sequence

Overview. Turns a one-line brief into a complete sequence across LinkedIn and email, with three distinct angles and every message written out ready to copy or push into a campaign.

Key features. Three angle variants per brief, day-by-day step planning across channels, copy for every step including connection notes and break-ups, and direct setup into the platform.

Pros and cons. The copy rules are encoded from real campaigns rather than generic best practice, and three angles make testing the default rather than an afterthought. It writes for LinkedIn and email specifically, so teams running calls or ads as primary channels will use only part of it.

Pricing. Free and MIT licensed. Pushing the sequence live needs a La Growth Machine account.

Best for. Anyone who writes sequences by hand and wants three testable angles instead of one.

→ Multichannel Campaign Builder

7. email-deliverability-audit (coldoutboundskills), best for protecting the inbox

Overview. Audits the technical and behavioural health of a cold email setup, covering authentication records, domain reputation, sending patterns and the content signals that route mail to spam.

Key features. SPF, DKIM and DMARC checks, domain and inbox warm-up review, sending volume patterns, plus a spam word checker and an incident response routine for when placement collapses.

Pros and cons. The only skill here that treats deliverability as an operational discipline with an incident path rather than a checklist. It is email-only, so it does nothing for LinkedIn-led motions.

Pricing. Free, MIT licensed, no paid dependency.

Best for. Teams sending cold email at volume who have seen reply rates drop without a copy change.

8. Reply Draft Assistant (La Growth Machine), best for triaging replies at volume

Overview. Triages every reply across active campaigns, classifies each by intent, and drafts an on-brand response ready to approve before it sends.

Key features. Cross-campaign reply triage, intent classification, on-brand draft generation, and approval before send so nothing leaves without a human seeing it.

Pros and cons. It solves the failure mode that kills most outbound programmes, which is replies going cold because nobody had time. It needs a connected account, and drafts still need reading: the approval step is deliberate, not a limitation to engineer around.

Pricing. Free and MIT licensed. Requires a La Growth Machine account to read conversations and send.

Best for. Teams past roughly fifty replies a week, where triage has become the bottleneck.

→ Reply Draft Assistant

9. daily-lead-steward (Zapier gtm-cheat-codes), best for inbound lead hygiene

Overview. A daily routine that reviews inbound leads for routing errors, missed follow-up and ownership gaps, then lists the leads that fell through the cracks with the reason each one did.

Key features. Daily inbound sweep, routing and ownership checks, missed-treatment detection, and a companion audit skill that looks back across a longer window.

Pros and cons. Brand-backed by Zapier and part of a well-organised set spanning GTM ops, sales and RevOps, legal and support. It assumes a Zapier-centred stack, so the value drops sharply if your routing lives elsewhere.

Pricing. Free, MIT licensed. Assumes a Zapier account for the workflows it references.

Best for. RevOps teams with meaningful inbound volume and a Zapier-based routing layer.

10. Campaign Impact Analyzer (La Growth Machine), best for ranking by CRM pipeline

Overview. Ranks campaigns by the pipeline they actually generated, by cross-referencing campaign data with deals in your CRM, and returns a Scale, Keep or Stop verdict for each.

Key features. Campaign-to-deal cross-referencing, ranking by pipeline rather than reply rate, per-campaign verdicts, and a breakdown by channel and by sequence step.

Pros and cons. It moves the optimisation target from reply rate to revenue, which changes which campaigns survive. Deal values come from your CRM, so the output is only as good as your CRM hygiene, and it needs both systems connected.

Pricing. Free and MIT licensed. Requires a La Growth Machine account and a connected CRM.

Best for. Heads of sales and RevOps who need to defend or kill campaigns with pipeline numbers.

→ Campaign Impact Analyzer

Sourcing and list building

find-lookalikes (YALC GTM-OS) expands a seed set of accounts into a scored lookalike list using firmographic and signal overlap. Best for teams that already know which accounts convert and want more of them.

list-building (ColdIQ) is a master skill covering list construction end to end, from source selection through segmentation. Best for SDRs building lists from scratch each week, if a ColdIQ key is acceptable.

Audience ICP Filter (La Growth Machine) filters an audience or lead list against your ICP and splits it into ready-to-sequence segments. Best for teams sitting on a large unsegmented list. → Audience ICP Filter

google-maps-list-builder (coldoutboundskills) builds local business lists from map data with address, category and contact fields. Best for local services and agency prospecting where LinkedIn coverage is thin.

Enrichment and data quality

enrich-with-signals (YALC GTM-OS) layers intent and trigger data onto an existing list rather than sourcing a new one. Best for teams whose list is fine but whose timing is not.

lead-enrichment (agent-gtm-skills) documents ICP scoring waterfalls and how to sequence providers so you stop paying twice for the same record. Best for anyone designing an enrichment waterfall, accepting that the repository is frozen.

list-quality-scorecard (coldoutboundskills) scores a list for deliverability and fit risk before you send to it, flagging the records likely to bounce or complain. Best for teams importing bought or scraped lists.

Copywriting and content

cold-email (ColdIQ) is the master copy skill: frameworks, sequence structure, subject lines and variation generation, with a large set of template sub-skills underneath. Best for teams whose cold email reads like every other cold email.

linkedin-content (ColdIQ) covers the organic side, from hooks and formats through to repurposing and scheduling. Best for founders and reps building a LinkedIn presence alongside outbound.

Post to Campaign (La Growth Machine) turns a LinkedIn post into a campaign by scraping its likers and commenters into an audience and filling a sequence calibrated to the post. Best for anyone whose posts get engagement that goes nowhere. → Post to Campaign

Campaign Challenger (La Growth Machine) benchmarks a new sequence against the campaigns that already performed, returns a verdict and a twelve-point score, and names the three fixes that matter. Best for the hour before you press launch. → Campaign Challenger

Reporting, RevOps and GTM ops

Team Performance Dashboard (La Growth Machine) turns outreach data into a per-rep cockpit and surfaces the practices behind top campaigns so they can be copied across the team. Best for managers with more than three reps. → Team Performance Dashboard

Weekly Performance Advisor (La Growth Machine) produces a two-tab weekly cockpit: what to do this week, and how each campaign is really performing. Best for the Monday review. → Weekly Performance Advisor

Objection Analyzer (La Growth Machine) ranks the objections your conversations actually produce, grades how the team answered them, and writes battle cards from your own threads. Best for enablement leads who want battle cards built from evidence. → Objection Analyzer

revops (marketingskills) covers RevOps diagnostics, forecasting and pipeline metrics, inside the largest general marketing skill library available. Best for RevOps teams wanting diagnostics not tied to a vendor.

For outbound sales teams and SDRs

Start with Sales Nav Search Builder to build the list, Multichannel Campaign Builder to write the sequence, and Reply Draft Assistant to handle what comes back. That is the full loop. Add cold-email when the copy needs a rethink, and signal-sourcer when static lists stop replying.

For RevOps teams

The RevOps set is different: revops for diagnostics and forecasting, daily-lead-steward for inbound routing hygiene, and Campaign Impact Analyzer to tie campaigns back to CRM pipeline. Add Audience ICP Filter when segmentation is the bottleneck, and Team Performance Dashboard when the question becomes which rep rather than which campaign.

For growth and demand generation

Growth teams get more from the upstream skills. positioning-icp first, because most growth experiments fail on positioning rather than execution, then Won-Deal ICP Finder to check it against closed-won evidence, then Post to Campaign to turn organic engagement into pipeline.

For GTM engineers building their own

Engineers want the pipeline, not the packaged output. prospect-discovery-pipeline and enrich-with-signals are the most forkable things here, meaning the parts you swap are provider keys, the MCP endpoint and the frontmatter description, with the reasoning left intact. lead-enrichment documents the waterfall logic well. For the terminal-native side of this stack, see our list of the best Claude Code skills for GTM engineers.

Method 1: Claude Code and other CLI agents

A skill is a directory containing a SKILL.md file. Copy the folder into ~/.claude/skills/ for personal use or .claude/skills/ inside a project, and Claude discovers it automatically. Cloning a whole library into that directory works the same way. Libraries packaged as plugins install through /plugin marketplace add and /plugin install instead, and /skills lists what is currently loaded.

The La Growth Machine suite installs with one command:

npx skills add lagrowthmachine/gtm-system

Method 2: Claude.ai and Claude Desktop

Enable skills under Settings, then Capabilities, and upload a SKILL.md there. This needs a Pro, Max, Team or Enterprise plan, since skills depend on code execution, and is not available on the free plan. Team and Enterprise administrators can provision skills organisation-wide, and skills enabled on Claude.ai also appear in Claude Desktop.

One gotcha: Claude Code supports frontmatter fields that Claude.ai does not, and a SKILL.md using them fails validation on upload. If a skill written for the CLI refuses to load in the browser, that is usually why.

What are the best Claude Skills for GTM?

For execution, meaning campaigns, replies and pipeline reporting, La Growth Machine’s eleven skills are the most complete set that both writes and runs the campaign. For sourcing and qualification, YALC’s GTM-OS. For signals and cold copy, ColdIQ. For positioning, positioning-icp.

Are Claude Skills for GTM free?

The skill files are free in every library here. What varies is the licence and the dependency: five of the seven repositories carry an MIT licence, two carry none, and several skills default to their publisher’s paid product, which is where the real cost appears.

Which Claude plan do I need to run skills?

In Claude Code, any plan works, since skills are folders the client reads locally. On Claude.ai and Claude Desktop they require Pro, Max, Team or Enterprise, because they depend on code execution, and are not available on the free plan.

Can I combine skills from different libraries?

Yes, and it is the normal setup. Claude chains skills through the conversation rather than through a schema, so the handoff is whatever the first skill wrote, not a structured record. The thing to watch is overlapping skills claiming the same trigger, which is what actually breaks a large install.

Do Claude Skills replace my outreach tool?

No. A skill is a set of instructions, not an execution engine. It writes the sequence, but sending it, rotating inboxes, respecting LinkedIn limits and recording the reply still happen in a platform. The same distinction applies to AI agents for sales automation: the agent decides, the platform executes.

What is the difference between a Claude Skill and an MCP server?

A skill is instructions, an MCP server is a connection. The skill tells Claude what to do, in what order, and what a good result looks like. The server lets Claude read and write in an external system. Most useful GTM setups need both.