Custom skills running across SEO, PPC, RevOps, and product marketing inside 250+ B2B SaaS programs. The same audits, briefs, and reports your team writes every week, configured once, repeated forever.
Examples from live programs, grouped by function. Every skill below is configured around your brand voice, your data, and your stack via MCP. Read what a production Claude skill is if you're new to the format.
Crawls the site, scores Core Web Vitals, flags index bloat, and groups fixes by engineering effort.
Five-subagent score for AI search readiness across citability, presence, technical, content, and schema.
Pulls top 10 results per query, classifies intent, and outputs the angle you can win on.
Brief with live SERP data, entity coverage, target FAQs, and brand voice baked in. 60 minutes to 10.
Live GAQL reads with anomaly detection, wasted-spend ranking, and the next best three actions.
Tracks CTR decay by creative cohort and flags the exact moment a hook stops working.
Reads 90 days of search terms, clusters waste, and writes the negative list with reason codes.
Daily skill run that compares spend to plan, flags drift, and emails the media lead before midnight.
Reads firmographic and behavioral signals, scores accounts against your ICP, and writes the rationale.
Flags duplicates, missing fields, and stale ownership. Outputs a queue your ops team can action.
Weekly deck on stuck deals, ageing opps, and next-step gaps. Lands in your inbox before Monday.
Pulls funnel data, isolates the drop, and writes the hypothesis your demand lead can test.
Pulls G2, Capterra, and competitor sites. Outputs positioning gaps and the talking points to win them.
Synthesizes closed-won data, support tickets, and interviews into a structured ICP your team agrees on.
Tests your statement with three buyer personas and notes their objections.
One-pager, sales deck, email sequence, social copy. End-to-end launch kit from one feature brief.
The marketing-skills community has shipped faster than any single team can. Start here before you build. These are the libraries we benchmark against and contribute back to.
Through MCP. Each tool exposes a connector. Your skill queries the connector at runtime, so the audit, brief, or report always runs against fresh data rather than a stale copy.
Through MCP. Each tool exposes a connector. Your skill queries the connector at runtime, so the audit, brief, or report always runs against fresh data rather than a stale copy.
Through MCP. Each tool exposes a connector. Your skill queries the connector at runtime, so the audit, brief, or report always runs against fresh data rather than a stale copy.
Through MCP. Each tool exposes a connector. Your skill queries the connector at runtime, so the audit, brief, or report always runs against fresh data rather than a stale copy.
Through MCP. Each tool exposes a connector. Your skill queries the connector at runtime, so the audit, brief, or report always runs against fresh data rather than a stale copy.
Through MCP. Each tool exposes a connector. Your skill queries the connector at runtime, so the audit, brief, or report always runs against fresh data rather than a stale copy.
Through MCP. Each tool exposes a connector. Your skill queries the connector at runtime, so the audit, brief, or report always runs against fresh data rather than a stale copy.
Through MCP. Each tool exposes a connector. Your skill queries the connector at runtime, so the audit, brief, or report always runs against fresh data rather than a stale copy.