AI Draft Revision Lane Cards

Purpose

Use this skill when a user has one or more AI-generated drafts and needs a visible review board before the text is sent, submitted, posted, or reused. The output is a printable or digital set of lane cards, status stickers, fact-check flags, tone-review labels, approval tabs, and a one-page handoff checklist.

This is a prompt-only organization workflow. It does not verify facts automatically, send messages, submit work, create hidden citations, or treat AI output as authoritative.

Use This Skill When

Use this skill when the user asks to:

• Review AI drafts for emails, essays, proposals, posts, reports, announcements, or client-facing notes.
• Separate raw AI output from human-edited and approved copy.
• Track fact checks, tone fixes, owner review, and final approval.
• Build a visible board for printed pages, notes, documents, slides, email drafts, chat outputs, or shared workspaces.
• Create labels or cards for a safe AI writing review workflow.

Do not use it to bypass school, workplace, legal, editorial, compliance, or publication rules.

Best Inputs

Ask only for the details needed to make useful cards. If the user does not know an answer, proceed with generic labels and placeholders.

• Draft type, such as email, article, proposal, school essay, social post, report, announcement, or client-facing note.
• Current draft location, such as printed pages, notes app, document, slide deck, email draft, chat output, or shared workspace.
• Reviewers or owners who must approve the draft.
• Known risks, such as facts, dates, figures, quotes, citations, promises, sensitive details, or brand voice.
• Preferred format, such as printable cards, digital board labels, sticky-note text, or a one-page checklist.

Do not ask for unnecessary private data, credentials, account access, confidential records, or full sensitive documents when a short excerpt would work.

Workflow

1. Identify the draft type, audience, current location, and external-use risk.
2. Create lane cards for raw AI output, keep, rewrite, verify, source needed, tone check, owner review, final, and archive.
3. Generate status stickers for strong sentence, unclear claim, unsupported number, sensitive detail, brand voice issue, duplicate idea, and final wording.
4. Build fact-check flags for names, dates, figures, quotes, promises, citations, instructions, and external-facing claims.
5. Add tone-review labels for too formal, too casual, too vague, too confident, too long, off brand, and needs human voice.
6. Create approval tabs with reviewer, decision, required change, deadline, and final send location.
7. Provide a one-page AI draft handoff checklist that makes human accountability visible before any external use.

Output Format

Return the artifact set in this order:

1. Draft Board Snapshot

• Draft type:
• Current location:
• Intended audience:
• External use risk:
• Review owner:
• Assumptions:

2. Revision Lane Cards

Provide copy-ready card text for each lane:

• Raw AI Output
• Keep
• Rewrite
• Verify
• Source Needed
• Tone Check
• Owner Review
• Final
• Archive

3. Draft Status Stickers

List short stickers with one-line meanings and suggested placement.

4. Fact-Check Flags

Create flags for names, dates, figures, quotes, promises, citations, instructions, and external-facing claims. Include a field for source or reviewer confirmation.

5. Tone-Review Labels

Create labels that help a reviewer mark tone issues without rewriting the whole draft immediately.

6. Approval Tabs

Include fields for reviewer, decision, required change, deadline, final send location, and approval date.

7. One-Page Handoff Checklist

Include steps for separating raw output, preserving user-provided facts, checking claims, removing sensitive details, confirming tone, recording approval, and placing final copy in the correct destination.

Style Rules

• Keep each card short enough to fit on a sticky note, index card, label, or digital board column.
• Use visible status words that make the next action clear.
• Use placeholders such as [REVIEWER], [DEADLINE], [SOURCE], and [FINAL LOCATION].
• Keep raw AI text visually separate from final approved wording.
• Make uncertainty visible instead of smoothing it away.
• Prefer plain labels over long explanations.

Safety Boundary

• Do not verify facts automatically, fabricate sources, generate hidden citations, or invent evidence.
• Do not send, post, submit, approve, or publish anything.
• Do not bypass school, workplace, client, editorial, legal, or compliance rules.
• Do not treat AI-generated text as final without explicit human review.
• Do not ask for credentials, private keys, account access, or unnecessary sensitive data.
• Flag names, dates, figures, quotes, promises, citations, and sensitive details for human confirmation.

Example Prompts

• "Make lane cards for reviewing this AI email draft before I send it."
• "I need an AI writing revision board for a proposal draft."
• "Create fact-check flags and approval tabs for several AI-generated social posts."
• "Turn this chat output into a safe draft review checklist."