← ClaudeAtlas

content-quality-gatelisted

A nine-check pass or fail gate that every piece of buyer-facing content must clear before it ships, covering audience specificity, problem-first framing, the belief it advances, the method it reflects, the shift it shows, terminology fidelity, sourced proof, human voice, and a single call to action. Use this skill whenever the user asks to draft or review a LinkedIn post, email, blog, proposal, deck, landing page, one-pager, or any external content; whenever they ask whether a draft is ready to ship, on brand, or good enough; and as the final check after any other skill has produced a draft. Trigger on "QA this", "is this ready", "does this pass", "brand review", and "make this sharper".
ingridleiria/breaktalk-skills · ★ 0 · AI & Automation · score 72
Install: claude install-skill ingridleiria/breaktalk-skills
# Content Quality Gate Most weak content is not badly written. It is written to nobody, about nothing anyone is losing sleep over, with numbers no one can trace. This skill is the gate that sits above every format skill: another skill produces the draft, this one decides whether it ships. Nine checks, each with a literal pass test. Any red is a rewrite, not a discussion. Run it in one of two modes. In drafting mode, resolve checks 1, 3, 4, and 7 before writing a word. In review mode, read the full piece once, then go check by check and quote the specific sentence that passes or fails each one. Never answer "looks good"; apply the test. ## First, load the standard This gate is generic by design. Before the first run for a given organization, capture its specifics once and reuse them: - The buyer segments and personas the organization sells to, with the pains each feels in their own words. - The two or three core beliefs the organization argues for. - The named method, process, or framework the content must reflect, with its real stage or component names. - The terminology rules: what each proprietary term is and is not, what is capitalized, what is never said. - The approved claims list: every statistic and client result cleared for external use, with its source. - The house voice rules and the default call to action. Store these in a `standard.md` beside this skill. Without them, checks 1, 3, 4, 6, and 7 can only be checked in outline. ## The nine checks **1. Written