review

name	review
description	Review a blog post draft for readability, interestingness, and factuality. Use when Tim asks to "review", "critique", "check", or "give feedback" on a post. Applies Tim's writing style standards and pushes for provocative, scannable, memorable writing.
version	1.0.0

Blog Post Review

Review blog post drafts against three priorities, in order. Produce a structured review — not a rewrite. Flag specific problems with line-level references and suggest fixes inline.

Priority 1: Readability (Scannability)

The reader should be able to half-read the post — eyes skimming, not deeply parsing — and walk away with basically the right point. This is the highest priority.

What to check:

Bold skeleton test. Read only the bold text in the post. Does it tell a coherent story? If someone read only the bold phrases, would they get the main argument? Flag sections where bold is missing, overused, or marking the wrong words.

Anchor paragraph test. Look for short (5-8 word) standalone paragraphs that act as visual speed bumps. These should appear every few paragraphs. If there's a long stretch without one, flag it. If they're weak or generic ("Let me explain"), suggest a punchier replacement.

Paragraph length. Any paragraph over 4 sentences is a problem. Flag it. Most paragraphs should be 1-2 sentences. If a paragraph has 3-4 sentences, check that it earns the length.

Section flow. Can you tell what each section is about from the heading alone? Do sections follow a logical progression? Flag any section that buries its point.

The squint test. Imagine squinting at the post so you can only see the shape of the text — short blocks, bold words, headings. Does it look scannable or does it look like a wall? Flag walls.

How to report readability issues:

For each issue, give:

• The location (section heading or paragraph number)
• What the problem is
• A specific suggestion (not "make it better" — show what "better" looks like)

Priority 2: Interestingness (The Hook)

The title and intro must pull the reader in. The mechanism: juxtapose modern-day intuition. Say the thing the reader already half-believes but hasn't heard said out loud. Then complicate it.

What to check:

Title. Is it provocative? Does it create tension or curiosity? A good title makes someone think "wait, what?" or "finally someone said it." A bad title describes the content. Flag descriptive titles.

Opening line(s). The first 1-3 paragraphs need to land hard. They should target the reader's unspoken thoughts — something they feel but haven't articulated. Check:

• Does the opener feel like it was written for the target reader?
• Does it create a "yes, but..." reaction?
• Is it bold enough? Or is it throat-clearing?

The intuition gap. Every post should push against some modern-day intuition. If you can't identify what intuition the post is challenging, that's a problem. Ask Tim what he thinks the modern-day intuition is so you can evaluate how hard the post pushes against it. This is not optional — ask.

Intro minimalism. The intro should orient the reader in 1-3 paragraphs and drop them in. If it takes more than 3 paragraphs before the post gets interesting, flag everything before that as cuttable.

Analogies. Are there analogies bridging from familiar territory to the new idea? Are they fresh or cliched? Do they earn their space?

Thesis gravity. Every section should pull back toward the post's main thesis. If a section wanders into supporting material (personal setup, technical detail, background), check whether it ties back. Short callout paragraphs at the end of sections are especially prone to floating — they should bridge the local point back to the central argument, not just label what happened. If a section could appear in a different post without changes, it's not tied tightly enough.

How to report interestingness issues:

• For title: suggest 2-3 alternatives that are more provocative
• For openers: show what a punchier version could look like
• For the intuition gap: state what you think the post is pushing against and ask Tim if that's right

Priority 3: Factuality (Armor, Not Accuracy)

This is NOT a fact-check. The goal is not to make the post more accurate. The goal is to make strong points stick — like an earworm. Memorable beats factual. But: anticipate where a reader will push back and preemptively address it, without hedging so much that the point dissolves.

What to check:

Vulnerable claims. Identify claims that a knowledgeable reader would immediately object to. For each:

• State the likely objection
• Suggest a one-sentence steelman or preemptive acknowledgment
• Do NOT suggest removing the claim or softening it. Instead, suggest how to hold the claim while acknowledging the counterpoint.

Over-hedging. Flag any hedging language ("it might be", "perhaps", "arguably", "it's possible that") and ask: is this hedge load-bearing (acknowledging real uncertainty) or is it just throat-clearing? If throat-clearing, suggest cutting it.

Under-hedging. Flag any claim stated with absolute certainty that is easily disprovable. Suggest a way to hold the strong opinion while adding a crack of daylight ("This is almost always true — the exception being...").

Stickiness. For the 2-3 most important points in the post, evaluate: will this stick in someone's brain? Is it phrased memorably? If a point is true but forgettable, suggest a rephrasing that's more vivid, concrete, or surprising.

How to report factuality issues:

For each issue:

• Quote the specific text
• State the likely reader objection or the stickiness problem
• Suggest a specific fix that keeps the opinion strong

Review Format

Structure the review as:

## Readability

[Issues organized by severity]

## Interestingness

**What intuition is this post pushing against?**
[Your best guess — then ask Tim to confirm or correct]

[Title and opener feedback]
[Other interestingness issues]

## Factuality

[Vulnerable claims with suggested armor]
[Hedging issues]
[Stickiness suggestions for key points]

## Top 3

The three most impactful changes, in order:
1. ...
2. ...
3. ...

What NOT to do

• Don't rewrite the post. Give targeted, specific feedback.
• Don't suggest adding summaries, conclusions, or "in this post we'll discuss..." framing.
• Don't flag style issues that match Tim's voice (fragments, starting with "But", casual abbreviations, internet-isms).
• Don't suggest making claims weaker. Suggest making them stronger and better armored.
• Don't suggest adding emojis, bullet-point summaries, or TL;DR sections.
• Don't praise what works unless it's genuinely surprising or non-obvious — Tim doesn't need validation, he needs useful critique.
• Don't apply generic "good writing" advice. Apply Tim's writing standards from the writing-style skill.