What it is.
Written by one of our co-founders, this is the inside account of how TripleDart builds and runs Claude skills across client work.
It separates a prompt from a skill, then documents the categories we deploy: brief generation, keyword research, internal and external linking, GEO audits, re-optimization, fact checking and ICP scoring.
The numbers make it concrete. Brief production moved from 45 to 60 minutes down to under 10, at roughly $0.10 per brief run, or $3 to $5 for a batch of 50.
What you get.
- The distinction between a prompt and a skill, with a worked comparison
- The structure we use to write one, and how skills get organized
- Six skill categories running in production, with what each one covers
- Named examples including a 25-node content brief pipeline and a GEO audit skill
- The scoring matrix for deciding what deserves a skill, with the qualifying threshold
- Cost and time figures: under 10 minutes per brief, about $0.10 a run
Read it at the source.
Read the guide
Are these skills hypothetical?
Do I need to hire you to use them?
What is the difference between a prompt and a Claude skill?
How much time does a skill save on a content brief?
What marketing work is worth turning into a skill?
What does running AI marketing workflows cost?