Critique a frozen writing artifact against an explicit rubric for purpose, structure, evidence, consistency, clarity, and voice without editing it.
Scanned 9/2/2026
Install to Claude Code
npx -y skills add kinqsradiollc/BrainRouter --skill writing-critique-skill --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Writing Critique Skill?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/kinqsradiollc-writing-critique-skill)More formats (shields.io, HTML) on the badges page.
---
name: writing-critique-skill
description: Critique a frozen writing artifact against an explicit rubric for purpose, structure, evidence, consistency, clarity, and voice without editing it.
allowed-tools: [read_file, list_dir, grep_search, glob_files, knowledge_list, knowledge_search, fetch_url, web_search, research_note]
---
# Writing critique
## Overview
Review one frozen artifact version without taking over authorship. Findings trace
to the supplied brief, rubric, and evidence, then give the primary writer bounded
options while preserving author intent and decision ownership.
## When to Use
Use after audience, purpose, intended effect, non-negotiable meaning, factual
constraints, artifact version, and critique rubric are explicit. Do not use to
silently rewrite, polish, or replace the artifact.
## Workflow
1. Confirm the frozen version, audience, purpose, intended effect, constraints,
non-negotiable meaning, voice, and rubric. Name missing review inputs.
2. Read the artifact continuously once, then inspect it in ordered passes:
purpose and completeness; structure and pacing; evidence and consistency;
paragraph logic; clarity and voice; mechanics.
3. Verify material factual or source-dependent claims against permitted sources
when needed. Never treat an unsupported claim as correct because it is fluent.
4. Classify findings as blocker, major, minor, or advisory. For each, report the
precise location, violated criterion, impact, evidence, and bounded options.
5. Separate confirmed defects from preference and uncertainty. Flag feedback
that would alter meaning, emphasis, or voice for the author's decision.
6. End with rubric coverage, the highest-impact revision order, and unresolved
questions. Return no rewritten artifact.
## Verification
- [ ] Every finding maps to a supplied criterion and precise artifact location.
- [ ] Structural and evidence issues precede sentence-level preferences.
- [ ] Factual checks retain source provenance and uncertainty.
- [ ] The output preserves author ownership and contains no silent rewrite.
## Red Flags
- Flattening a distinctive voice into generic professional prose.
- Listing preferences as defects without connecting them to audience or purpose.
- Suggesting sentence polish while a blocking structural or evidence gap remains.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!