doc-review

Orchestrates multi-perspective review of technical documents using independent reviewer subagents.

Updated Jun 24, 2026
One-click install
npx skills add https://github.com/kashman001/ai-workspace-template --skill doc-review-kashman001
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: doc-review
Source: https://github.com/kashman001/ai-workspace-template/tree/main/skills/doc-review
Command: npx skills add https://github.com/kashman001/ai-workspace-template --skill doc-review-kashman001

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Technical documents often fail their readers because authors cannot see their own blind spots: mixed content modes, wrong audience assumptions, and untested instructions. This Skill runs a structured, multi-perspective review so you get prioritized, actionable findings instead of vague feedback. ## Core Features & Use Cases - Audience Gate First: Establishes audience segments, jobs-to-be-done, and success criteria before any review begins, so findings are judged against real readers. - Six Independent Reviewers: Dispatches specialist subagents covering structure (Diátaxis), per-segment audience fit, task walkthrough, language clarity, accuracy, and skim/forward testing — each isolated from the others. - Synthesized Output: Merges findings into a verdict, root-cause diagnosis, severity-ranked findings, an audience coverage matrix, and a recommended restructure. - Use Case: Before publishing an internal platform guide, run this Skill to discover that your tutorial buries architecture rationale mid-procedure and that one audience segment cannot complete the setup steps. ## Quick Start Ask the AI to run the doc-review skill on your document, for example: review docs/onboarding-guide.md with the doc-review skill.

Frequently Asked Questions about doc-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review a technical document for multiple audiences?▼

Define each audience segment with its role, technical depth, and job to be done, then evaluate the document once per segment. This Skill automates that by spawning one audience-fit reviewer per segment and merging their findings into a coverage matrix.

What is the Diátaxis framework for documentation review?▼

Diátaxis classifies documentation into four modes: tutorials, how-to guides, reference, and explanation. The review checks whether a document mixes these modes, since mode-mixing is a common cause of documents feeling simultaneously too detailed and too sparse.

Can the review run without parallel subagent support?▼

Yes. If the runtime cannot dispatch parallel subagents, each reviewer runs sequentially in an isolated context. Reviewers must never see each other's findings so their perspectives stay independent.

Does the document review rewrite my content automatically?▼

No. The Skill diagnoses problems and proposes a restructure but does not rewrite the document unless explicitly asked. It delivers a verdict, prioritized findings, and a recommended heading outline first.

Why does the review skip readability scores like Flesch-Kincaid?▼

Grade-level scores penalize necessary domain vocabulary and measure nothing meaningful for technical documents. The language reviewer instead flags specific issues: passive voice, long sentences, ambiguous pronouns, and inconsistent terminology.