Skill: QA Reviewer for AI Drafts
A review skill for catching weak claims, structural drift, tone problems, and hidden assumptions in AI-generated drafts before they go live.
Use cases
Content & Writing, Operations & Workflow, Development & Code
Platforms
Claude, GPT, Model-Agnostic
Jump to a section
Workspace-ready resource
Copy the raw material, then check the setup, output contract, and failure modes.
# QA Reviewer for AI Drafts Review standard:- Identify the actual issue, not just cosmetic edits.- Separate structural problems from polish.- Flag weak evidence, unsupported claims, repetition, and vague language.- Recommend the smallest useful fix where possible. Return:1. Issues found2. Severity3. Why it matters4. Suggested fix5. Whether a human must review before use Do not rewrite everything unless the draft is fundamentally broken.Workspace translation
Turn this resource into an inspectable run.
Best next step
Use it to define the quality bar before pairing it with a playbook.
From resource to system
Use it once
Use the skill as a capability pack for a specific job.
Make it reusable
Adapt the role, standards, tone, and review rules into a reusable operator.
Decide after the run
Use it to define the quality bar before pairing it with a playbook.
Quality bar
Before using this, check the contract.
What input does this require?
What output should it produce?
Where can it fail?
What should a human review?
Recommended path
When to Use This
Use this when you already have AI output but want a disciplined second layer before it gets published, sent, or pushed into a workflow.
It is useful across copy, docs, internal memos, support replies, and agent outputs that need a review step without full manual rewriting.
Why It Works
The skill is designed to behave like a reviewer, not a co-writer. That means diagnosis first, severity second, and only then proposed changes.
The “smallest useful fix” instruction is important because review systems often create churn by rewriting everything instead of identifying the real fault line.
How to Customise
Add domain checks if you need them: legal risk, citation quality, policy alignment, code correctness, or brand voice.
You can also define a tighter severity rubric if the output feeds a production workflow.
Limitations
This is still a model reviewing model output. It improves quality control but does not replace a human for high-stakes review.
It also depends on having a clear standard. If “good” is vague, the review quality will be vague too.
Model Notes
Claude is strong at candid critique without overreacting to minor issues.
GPT works well when the severity scale is defined tightly. Model-agnostic overall.
Related Resources
Browse SkillsPrompt: Self-Evaluation Checklist
A finishing prompt that makes the model critique its own draft for clarity, evidence, tone, and structural weak points.
Pattern / playbook seed
Content & Writing · Development & Code
Framework: Prompt Audit Checklist
A 15-point checklist for evaluating any prompt before putting it into production. Catches the most common prompt failures: vague instructions, missing constraints, absent error handling, and untested edge cases.
Knowledge / rubric seed
Operations & Workflow · Strategy & Planning
Framework: Human-in-the-Loop Review Design
A framework for deciding where human review should sit in an AI workflow, what must be checked, and what can safely move faster.
Knowledge / rubric seed
Operations & Workflow · Strategy & Planning
Related Guides
Human Review Is Not a Failure Mode
A practical argument for treating human review as part of intelligent workflow design rather than as evidence that the AI system failed.
Prompt Testing: How to Know If Your Prompt Is Good
A practical guide to prompt evaluation that goes beyond vibes and looks at repeatability, failure cases, and revision discipline.
Need this operationalized?
Turn the pattern into a workspace system.
Use MPV for the private workspace loop, or work with Encanta to turn operators, playbooks, context, and review flows into a team-ready implementation.