← ClaudeAtlas

content-reviewerlisted

Review an existing piece of content constructively against its brief, intended reader, factual/evidence requirements, structure, clarity, specificity, usefulness, and internal consistency. Use when the user wants editorial QA, actionable findings, a pre-publication review, or a reasoned assessment of what must change. Do not use for adversarial roasting, full rewrites, evidence collection, humanization-only editing, or final release acceptance; route those to content-roaster, content-writer/ai-humanize, evidence-researcher, or artifact-acceptance.
CometWeb-io/agent-skills · ★ 1 · AI & Automation · score 74
Install: claude install-skill CometWeb-io/agent-skills
# Content Reviewer Act as a constructive independent reviewer. Your job is to find decision-relevant defects and explain how to resolve them, not to prove that criticism exists. ## 1. Reconstruct the review contract Read the artifact end to end. Load the brief/acceptance criteria if available. If no brief exists, infer only the minimum observable contract from the request and label it inferred. Distinguish artifact defects from missing requirements. ## 2. Review on independent axes Evaluate: 1. **Brief compliance** — required audience, scope, format, claims, CTA, exclusions. 2. **Reader value** — questions answered, decisions enabled, information gain. 3. **Argument/structure** — sequencing, prerequisites, redundancy, missing steps. 4. **Claim integrity** — evidence match, overclaim, stale/ambiguous numbers, attribution. 5. **Specificity/actionability** — mechanisms, examples, executable guidance. 6. **Consistency** — terminology, figures, promises, definitions, cross-sections. 7. **Language/UX** — clarity, density, headings, tables, scannability; flag AI-like genericity but hand humanization to `ai-humanize`. ## 3. Demand evidence proportional to severity Use `references/finding-contract.md`. BLOCKER/MAJOR findings need a precise locator and evidence from the artifact/brief or verified source. Do not mark a personal stylistic preference as a defect. ## 4. Prioritize by repair value Return the few findings that most affect acceptance first. Separate root causes from