---

name: review-all description: Review all blog articles in the content directories. Iterates through posts and in-research directories, reviewing each article one by one with fresh context. Supports --fix-all to auto-apply all changes across all articles.

Review All Posts Skill

Reviews all technical articles on sujeet.pro. Iterates through content directories and applies /review-posts to each article.

Note: This skill reviews content only. For code review, use /review-code.

Invocation

• /review-all - Review all posts
• /review-all --fix-all - Auto-fix all issues
• /review-all --fix-critical - Auto-fix critical only
• /review-all --dry-run - List articles without reviewing
• /review-all --category=web - Review specific category only

Flags

• --fix-all - Auto-apply all fixes to every article
• --fix-critical - Auto-apply critical fixes only
• --dry-run - List articles that would be reviewed
• --category=<name> - Only review specific category

Workflow

Copyflowchart TD
    A[User Request] --> B[Discover Articles]
    B --> C[Build Queue]
    C --> D{More Articles?}
    D -->|Yes| E[Clear Context]
    E --> F[Review Article]
    F --> G{Fix Flags?}
    G -->|--fix-all| H[Apply All Fixes]
    G -->|--fix-critical| I[Apply Critical Only]
    G -->|No flags| J[Report Only]
    H --> K[Update lastReviewedOn]
    I --> K
    J --> L[Add to Summary]
    K --> L
    L --> D
    D -->|No| M[Final Summary]

Phase 1: Discover Articles

Content Directories

Discovery

Copyfind content/posts content/in-research -name "*.md" -not -path "*/drafts/*"

Phase 2: Build Article Queue

Ordering

1. By last updated (oldest first) - prioritize stale content
2. By category - group related articles
3. By complexity - longer articles first

Display List

Copy## Articles to Review

Found **[N]** articles:

### Posts ([count])

1. [title] - `path` - Last updated: [date]
2. [title] - `path` - Last updated: [date]

### In-Research ([count])

1. [title] - `path` - Last updated: [date]

Proceed with review? [Y/n]

Phase 3: Review Each Article

Context Management

CRITICAL: Start fresh for each article:

Copy---
## Reviewing Article [N] of [Total]

**Article**: [title]
**Path**: [path]

> Starting fresh context for this review.
---

Comprehensive Review Checklist

For each article, execute ALL of the following checks:

3.1 Metadata Check

• lastReviewedOn present and recent
• Tags from content/tags.jsonc (NOT src/content/tags.json)
• Valid YAML frontmatter
• Date in folder name matches content

3.1.1 Tag Review (IMPORTANT)

1. Read content/tags.jsonc to get all valid tag IDs
2. Analyze post content to identify relevant topics
3. Add relevant tags that match the post content:
   • Use tag id values (e.g., web-performance, not Web Performance)
   • Add all applicable tags based on content topics
   • NEVER remove existing tags from the post
4. Add new tags to tags.jsonc if needed:
   • If a relevant topic has no matching tag, add it to content/tags.jsonc
   • Place new tag in appropriate category section
   • Follow existing format: { "id": "slug-format", "name": "Display Name" }
5. Validate all tags in the post exist in tags.jsonc

3.1.2 Title Review (IMPORTANT)

Review the H1 title for quality and appropriateness:

• Title accurately reflects the content
• Specific enough to distinguish from similar topics
• No clickbait or sensationalism
• Under 70 characters (SEO), but prioritize clarity
• Uses descriptive technical format

Update title if:

• Content has evolved beyond original title scope
• Title is too vague (e.g., "Tips" instead of specific topic)
• Title doesn't match what the article actually covers

3.1.3 Slug Review (IMPORTANT)

Review the folder slug (folder name containing the post):

• Slug reflects main topic
• Concise (3-5 words max)
• Lowercase with hyphens, no special characters
• Date prefix intact (YYYY-MM-DD)

Rename folder if:

• Content topic has significantly changed
• Slug is too vague or generic
• Slug doesn't match updated title

Renaming process:

1. Rename folder with new slug
2. Update any internal links referencing old path
3. Verify build succeeds

3.2 Structure Analysis

• Clear title (H1) - descriptive, not clickbait
• Abstract paragraph (2-4 sentences) - sets context
• Overview mermaid diagram - visualizes core concept
• Comprehensive TLDR section with subsections
• Proper section hierarchy (H2 → H3 → H4)
• References section present (REQUIRED)
• NO manual Table of Contents (auto-generated)

3.3 Content Inventory

Create inventory of:

• All claims made (with/without citations)
• All code examples (collapse usage, correctness)
• All diagrams (mermaid/ASCII/images)
• All external references (URLs, papers, docs)

3.4 Fact-Check (CRITICAL)

Verify Technical Claims:

1. Search official documentation for each claim
2. Cross-reference multiple authoritative sources
3. Flag discrepancies, outdated info, or speculation stated as fact
4. Ensure terminology is correct and current

Check Code Examples:

1. Syntax correct for stated language/version?
2. Logic correct and idiomatic?
3. Follows current best practices?
4. Error handling appropriate?
5. Boilerplate collapsed? (imports, setup, teardown)
6. Would a staff engineer find issues?

Validate References:

1. URLs accessible and not 404?
2. Content at URL actually supports the claim?
3. Better/more authoritative sources available?
4. Every significant claim has inline reference?

3.5 Code Block Review

Every code block must:

• Have title attribute for context
• Use collapse for any lines not essential to the key concept:
  • Imports and type definitions
  • Setup, configuration, and boilerplate
  • Middle sections (helpers between key functions)
  • Utilities and cleanup code
• Use multiple collapse ranges when needed: collapse={1-5, 12-18, 25-30}
• Use line highlighting {n-m} for key concepts within visible sections
• Be syntactically correct
• Follow idiomatic patterns for the language

Key principle: Only the lines demonstrating the key concept should be visible. Everything else should be collapsed but expandable.

3.6 TLDR Quality Check

• Main concept defined (1-2 sentences, not vague)
• 3-6 themed subsections covering key areas
• 3-6 bullet points per subsection
• Bold for key terms and concepts
• Standalone useful (not just a teaser)
• Includes key trade-offs
• A staff engineer could use as reference

3.7 Quality Assessment

Technical Accuracy (HIGHEST PRIORITY):

• All claims verifiable and verified
• Inline references present for claims
• No speculation presented as fact
• Correct, current terminology
• Code would pass PR review

Authoritative Tone:

• Assertive, not hedging unnecessarily
• Direct statements, not "might be" or "could possibly"
• Confident where evidence supports
• Honest about unknowns and limitations

Conciseness (NO FILLER):

• No padding or filler sentences
• No tutorial-style hand-holding
• No obvious statements ("security is important")
• Every paragraph earns its place
• No "In this article, we will..."
• Reading time < 30 minutes (60 max)

Completeness:

• Covers design reasoning (why, not just what)
• Documents assumptions and constraints
• Historical context where relevant
• Explicit pros/cons for all approaches
• Alternative approaches mentioned
• Real-world examples included
• Nothing presented as "the best solution"

Staff/Principal Engineer Standard:

• Could be cited as authoritative reference
• Handles edge cases and subtleties
• Discusses failure modes
• Performance implications noted
• Security considerations covered

Score each area (1-5):

3.8 Generate Report

Copy# Review Report: [Article Title]

## Summary

**Overall Quality**: [Excellent/Good/Needs Work/Major Issues]
**Technical Accuracy**: [Score/5]
**Staff Engineer Ready**: [Yes/No]
**Last Reviewed**: [Date]

## Critical Issues (Must Fix)

### Issue 1: [Title]

- **Location**: [Section/line]
- **Problem**: [Description]
- **Evidence**: [Research findings with source]
- **Fix**: [Specific recommended change]

## Fact-Check Results

| Claim   | Verified | Source | Notes   |
| ------- | -------- | ------ | ------- |
| [Claim] | ✓/✗      | [URL]  | [Notes] |

## Code Review Results

| Block   | File   | Issues   | Collapse Needed |
| ------- | ------ | -------- | --------------- |
| [Title] | [Line] | [Issues] | [Yes/No]        |

## Improvements Needed

### Structure

- [Issues]

### Missing Elements

- [ ] Item

### Outdated Information

- [Items needing updates]

## Anti-Patterns Found

- [ ] Tutorial-style content
- [ ] Filler statements
- [ ] Unsubstantiated claims
- [ ] Silver bullet thinking
- [ ] Migration timelines/development plans (should be removed unless explicitly requested)

## Recommendations

### High Priority (Critical)

1. [Fix]

### Medium Priority (Improvements)

1. [Enhancement]

### Low Priority (Polish)

1. [Polish item]

Handle Fixes

With --fix-all:

• Apply all fixes
• Update lastReviewedOn
• Run build validation at end

With --fix-critical:

• Apply critical fixes only
• Update lastReviewedOn

Without flags:

• Generate report
• No modifications

Track Progress

Copy## Review Progress

| #   | Article | Status      | Critical | Improvements | Polish  |
| --- | ------- | ----------- | -------- | ------------ | ------- |
| 1   | [title] | Done        | 2 fixed  | 5 fixed      | 3 fixed |
| 2   | [title] | In Progress | -        | -            | -       |
| 3   | [title] | Pending     | -        | -            | -       |

Phase 4: Final Summary

Copy# Full Site Review Summary

**Date**: YYYY-MM-DD
**Articles Reviewed**: [N]

## Overview

| Category    | Count   | Critical | Improvements | Polish  |
| ----------- | ------- | -------- | ------------ | ------- |
| Posts       | [N]     | [X]      | [Y]          | [Z]     |
| In-Research | [N]     | [X]      | [Y]          | [Z]     |
| **Total**   | **[N]** | **[X]**  | **[Y]**      | **[Z]** |

## Articles by Quality Score

### Excellent (5/5)

- [title] - `path`

### Good (4/5)

- [title] - `path`

### Needs Work (3/5)

- [title] - `path`

### Major Issues (1-2/5)

- [title] - `path`

## Common Issues Found

### Most Frequent Critical Issues

1. **[Issue]** - Found in [N] articles
2. **[Issue]** - Found in [N] articles

### Most Frequent Improvements

1. **[Issue]** - Found in [N] articles

## Articles Requiring Manual Attention

1. **[title]**
   - Issue: [description]
   - Action: [recommended]

## Actions Taken (with --fix-all)

- Critical issues fixed: [N]
- Improvements applied: [N]
- lastReviewedOn updated: [N] files

## Next Steps

1. Review "Major Issues" articles
2. Address manual-attention items
3. Run: `npm run build`
4. Validate: `npm run validate:build`

Batch Processing

Memory Efficiency

Process in batches of 5 to prevent context overflow:

CopyBatch 1 (Articles 1-5)
├── Review each
├── Apply fixes
└── Generate batch summary

[Clear detailed context, keep summary]

Batch 2 (Articles 6-10)
├── Review each
├── Apply fixes
└── Merge into cumulative summary

...

Final Summary

Error Handling

Per-Article Errors

If article fails:

1. Log error details
2. Mark as "Review Failed"
3. Continue with next article
4. Include in "Manual Attention"

Build Validation

After all fixes:

Copynpm run build
npm run validate:build

Quality Standards Reference

Content Quality

• Abstract sets context
• Overview diagram present
• TLDR comprehensive
• Inline references for claims
• Trade-offs discussed

Conciseness

• No padding/filler
• No tutorial-style
• Every paragraph earns place
• Reading time < 30 min

Code Blocks

• All non-essential lines collapsed (imports, setup, middle sections, helpers)
• Multiple collapse ranges used when needed: collapse={1-5, 12-18, 25-30}
• Titles present
• Key lines highlighted

Structure

• No manual ToC
• References section
• Proper hierarchy

Tags

• All tags valid (exist in content/tags.jsonc)
• Relevant tags added based on content
• New tags added to tags.jsonc if needed
• Existing tags NOT removed

Title & Slug

• Title accurately reflects content
• Title is specific and descriptive
• Slug matches content topic
• Slug is concise (3-5 words)

Example Session

Dry Run

CopyUser: /review-all --dry-run

Claude: ## Articles to Review (Dry Run)

Found **24** articles:

### Posts (15)
1. Node.js Event Loop - `content/posts/web/...`
2. HTTP Deep Dive - `content/posts/web/...`

### In-Research (9)
1. WebSocket Research - `content/in-research/...`

Run without --dry-run to start.

Full Review

CopyUser: /review-all --fix-all

Claude: ## Starting Full Site Review

Found **24** articles. Auto-fixing enabled.

---
### Reviewing Article 1 of 24
**Article**: Node.js Event Loop
**Path**: content/posts/web/2023-12-03-node-js/index.md

[Review process...]

**Fixes Applied**:
- Added 3 inline references
- Updated code blocks with collapse
- Added trade-off table
- Updated lastReviewedOn

---
### Reviewing Article 2 of 24
...

## Final Summary
[Complete report]

Internal Linking

When reviewing, verify all internal links use relative paths to .md files. This enables IDE navigation (Cmd+Click) and the rehype plugin transforms them to proper URLs at build time.

Correct format:

Copy[Link Text](../YYYY-MM-DD-slug.md)
[Link Text](../../category/YYYY-MM-DD-slug/index.md)

Examples:

Copy<!-- Correct - relative .md paths -->

[Web Performance Overview](../2025-03-03-wpo-overview.md)
[JavaScript Optimization](../2025-01-09-wpo-js.md)

<!-- Wrong formats (flag and fix) -->

[Wrong](/posts/deep-dives/web-fundamentals/wpo-overview) <!-- Direct URL, not IDE navigable -->
[Wrong](wpo-overview) <!-- Missing .md extension -->

Key rules:

• Use relative paths from current file to target .md file
• Include the full filename with date prefix
• The rehype plugin transforms these to /posts/<type>/<category>/<slug> URLs
• Enables Cmd+Click navigation in VS Code and other IDEs

During review: Convert direct URL links to relative .md paths where possible.

Reference Documents

IMPORTANT: Before reviewing, read these documents from the project root:

Usage: Use the Read tool with absolute paths (e.g., /path/to/project/llm_docs/content-guidelines.md) to read these files before starting work.

Tools Available

• Glob - Find articles
• Grep - Search content
• Read - Read articles
• Edit - Apply fixes
• Bash - Build validation
• WebSearch - Fact-check
• WebFetch - Verify references
• TodoWrite - Track progress