Review Blog Post

Perform a critical review of a blog post evaluating content, flow, and voice against the standards in voice.md, CLAUDE.md, and the patterns established across published posts.

Arguments

• •$1 - Post slug (e.g., the-layer-above) or file path. If omitted, review the most recently modified post in src/content/posts/.

Workflow

Step 1: Gather materials

1. •Read the target post fully
2. •Read voice.md (project root) - this is the single source of truth for voice, tone, punctuation, formatting, and style
3. •Read CLAUDE.md - specifically the Writing Style section and any project-specific conventions
4. •Read 2-3 most recent published posts (non-draft, by pubDate descending) for voice calibration. These are your reference for what "on-voice" sounds like in practice.

Step 2: Establish context

Before any evaluation, write a brief context section:

• •Thesis: What is the post's central argument or observation?
• •Prompt: What seems to have prompted this post (experience, trend, event)?
• •Arc: How does the post develop its idea (e.g., concrete opening -> analysis -> counterargument -> open close)?

This context frames the review. A post about personal experience has different content expectations than one making technical claims.

Step 3: Evaluate

Apply the criteria below across four dimensions. For each finding, assign a severity:

Be specific. Quote the problematic text and explain why it doesn't work. When suggesting changes, offer a concrete rewrite.

Step 4: Output

Use the output format defined at the bottom of this file.

Review Criteria

1. Content

Evaluate whether the post's argument is sound and well-supported.

Check for:

• •Thesis clarity: Can you state the central argument in one sentence? If not, the post may lack focus.
• •Evidence: Are claims supported by concrete evidence (personal experience, citations, examples)? Flag unsupported assertions.
• •Counterarguments: Does the post address obvious objections? Strong posts in this blog acknowledge tension rather than arguing past it.
• •Essentiality: Does every section serve the argument? Flag paragraphs that could be cut without losing anything.
• •Citations: Are external claims linked? Do links open in new tabs? (See voice.md for format.)

2. Flow

Evaluate how the post moves the reader through its ideas.

Check for:

• •Opening: Does it ground the reader in something concrete and specific? This blog never opens with abstractions or thesis statements. It opens with situations - a client email, a tennis club, a debugging session.
• •Arc: Does the post follow a natural progression? Typical pattern: concrete situation -> development and analysis -> complication or counterargument -> implications -> open close. Not every post follows this exactly, but there should be a discernible shape.
• •Transitions: Are section breaks (---) placed at genuine tonal or topical shifts? Do bridging sentences connect sections naturally rather than announcing what comes next?
• •Closing: See voice.md for closing style. Flag tidy conclusions, ceremonial callbacks to the opening, or unnecessary wrap-up paragraphs.
• •Pacing: Is paragraph length varied? Long analytical paragraphs should be balanced by shorter grounding ones. Flag monotonous paragraph lengths.

3. Voice

This is the most important dimension. voice.md is the authoritative reference - read it fully before evaluating. Compare against both voice.md guidelines and the reference posts you read.

Flag any violations of voice.md rules, including but not limited to: sentence rhythm, directness, personal grounding, tone, phrases to avoid, emphasis/formatting, and claims. When flagging, reference the specific voice.md section being violated.

4. Mechanics

Quick checks for formatting consistency per voice.md:

• •Punctuation: Per voice.md Punctuation section
• •Citations: Per voice.md Citations section
• •Frontmatter: Required fields present (title, description, pubDate). draft: true if not ready.

Output Format

## Review: [Post Title]

### Context
[Thesis, prompt, and arc as described in Step 2]

### Content
[Findings with severity tags]

### Flow
[Findings with severity tags]

### Voice
[Findings with severity tags]

### Mechanics
[Findings with severity tags]

### Summary
[2-3 sentences: overall assessment, most important thing to address, and what's working well]

Finding format

Each finding should follow this pattern:

[Severity] - [What the issue is]

[Quoted text from the post]

[Why it doesn't work, referencing the specific criterion]

[Suggested revision if applicable]