Juejin Article Optimizer (掘金技术文章优化助手)#

Take your technical draft from "just another post" to a high-ranking, high-engagement Juejin article. Analyze structure, readability, and SEO against platform best practices — then get a fully optimized version ready to publish.

Core Capabilities#

• Article diagnostic scoring: Rate your draft across 6 dimensions (title, structure, code/text ratio, readability, SEO, engagement hooks) with specific improvement targets
• Title optimization: Generate 3 high-CTR title variants based on Juejin's top-performing title patterns
• Structure reorganization: Recommend optimal heading hierarchy, paragraph length, and code/text balance for Juejin's mobile-first reading experience
• SEO enhancement: Keyword density analysis, internal link suggestions, tag optimization, and TOC anchor generation
• Code block improvement: Annotate code blocks with explanatory comments, add syntax highlighting language tags, suggest split/merge for readability
• Visual suggestion engine: Identify which sections would benefit from diagrams, screenshots, or GIFs — with placement recommendations
• Engagement boost: Add discussion-bait questions, practical takeaways, and community interaction hooks
• Publishing optimization: Tag strategy, best publish day/time, and cross-platform promotion snippets

Workflow (8 Steps)#

Step 1: Article Ingestion#

Input: User provides:

• Article draft: Markdown file, pasted text, or URL of existing Juejin article
• Goal: optimize-new (optimize before first publish) | rescue-existing (improve a low-performing published article) | translate-adapt (adapt English tech blog for Juejin)
• Target audience (optional): Beginner / Intermediate / Advanced

Output: Parsed article with structural metadata (word count, heading count, code block count, image count, reading time).

Step 2: Diagnostic Scoring#

Input: Parsed article. Action: Score on 6 dimensions (1-10 scale):

DimensionWhat We MeasureWeightTitle PowerCTR potential. Does it use Juejin-proven patterns (numbered lists, practical value, curiosity gap)?20%StructureHeading hierarchy, paragraph length, first-screen content value, section balance15%Code/Text RatioIs there enough explanation around code? Standard: 60% text, 40% code. Too much code → low readability.15%ReadabilitySentence length, terminology explanation, Chinese readability score, mobile readability20%SEO & DiscoverabilityKeyword density, tag relevance, title keyword placement, internal linking15%Engagement DesignDiscussion hooks, practical takeaways, call-to-action, comment-bait questions15%

Output: Radar chart (text/Mermaid) + overall score + breakdown. Comparison to Juejin's top-article benchmarks.

Step 3: Juejin Trend Analysis#

Input: Article topic + Juejin category. Action: Scrape Juejin's relevant category/channel for top-performing articles (last 30 days, >500 likes):

• Extract common title patterns (e.g., "一文搞懂...", "XXX最佳实践", "面试官问XXX")
• Analyze optimal article length (character count distribution of top articles)
• Code/text ratio benchmarks in the category
• Tag usage patterns (which tag combos get most exposure)
• Time-of-day posting patterns (extracted from publish times of trending articles)

Output: Category benchmark report with actionable patterns.

Step 4: Gap Analysis#

Input: Article diagnostic score + category benchmarks. Action: Compare article metrics against category benchmarks:

DimensionYour ArticleCategory AverageTop 10%GapTitle Power5/106.58.2🔴 -3.2Structure7/106.88.5≈Code/Text Ratio55% code35%28%🔴 Too much codeReadability6/107.28.8🟡 -2.8SEO4/106.08.0🔴 -4.0Engagement3/105.57.5🔴 -4.5

Output: Prioritized improvement list: "Top 3 fixes: 1. Add engagement hooks (gain +2.5) 2. Improve SEO (gain +2.5) 3. Optimize title (gain +2.0)."

Step 5: Title Optimization#

Input: Original title + category patterns. Action: Generate 3 optimized title variants using Juejin-proven patterns:

1. Numbered List Pattern: "XXX 的 5 个最佳实践，第 3 个最容易被忽略"
2. Practical Value Pattern: "用了 XXX 这么久，90% 的人不知道这个技巧"
3. Curiosity Gap Pattern: "为什么你的 XXX 总是慢 3 倍？一个被忽视的配置"

Output: 3 title variants with CTR potential scores and rationale. User selects or combines.

Step 6: Content Optimization#

Input: Original article + gap analysis + selected title. Action: Apply optimizations:

• Opening hook: Rewrite first 3 paragraphs — start with problem statement, data point, or relatable scenario (not "本文介绍了...")
• Structure adjustment: Ensure H2 level is scannable (reader can understand the article by reading only H2s). Add a TOC at top for articles >2000 chars.
• Code block enrichment: Add a "What this code does" comment before every code block. Add expected output example. Suggest language tag if missing.
• Paragraph trimming: Max 4 sentences per paragraph. Break walls of text. Add bold highlights to key sentences.
• Visual insertion points: Mark [💡 建议配图: XXX架构图] at 3-5 strategic locations.
• Practical takeaways: End each major section with "💡 要点总结" (1-2 sentences).

Output: Optimized markdown article. Diff available on request.

Step 7: SEO & Discoverability Enhancement#

Input: Optimized article. Action:

• Keyword optimization: Identify primary keyword (1) and secondary keywords (3-5). Ensure primary keyword appears in title, first paragraph, at least 2 H2s, and naturally 5-8 times in body.
• Tag strategy: Recommend 5 tags: 1 primary category tag + 2-3 topic tags + 1 traffic tag. Explain why each tag is chosen.
• TOC generation: Auto-generate linked table of contents for articles >5 H2 sections.
• Internal linking: Suggest 2-3 places to link to your own previous articles (if applicable).
• Description/summary: Write a 2-3 sentence article description for the Juejin feed preview.

Output: SEO optimization report + applied changes.

Step 8: Publishing Strategy#

Input: Optimized article + category benchmarks. Action: Generate publishing recommendations:

• Best publish time: Day of week + hour window based on category data (e.g., "周三 10:00–11:00 or 周四 14:00–15:00, 前端类文章最佳发布窗口")
• Cross-promotion snippets: 1 WeChat Moments share text, 1 tech community (V2EX/Zhihu) cross-post hook, 1 Twitter/X thread starter
• First-hour engagement plan: How to respond to first 5-10 comments to boost algorithm ranking (acknowledge, extend, ask follow-up)
• Series potential: If article is part of a broader topic, suggest a 3-article series plan to build follower momentum

Output: Publishing playbook with timeline + promotional copy.

Sample Prompts#

Prompt 1: Pre-Publish Optimization#

User: "我写了一篇关于Go并发编程的文章，帮我在发布前优化一下 [paste article]" Expected Output: Diagnostic score → title variants ×3 → optimized article with improved structure, annotated code blocks, visual suggestions, SEO enhancements, and publishing strategy.

Prompt 2: Low-Performing Article Rescue#

User: "这篇文章在掘金只有200阅读，帮我诊断为什么，然后优化 [URL: juejin.cn/post/...]" Expected Output: Diagnostic report identifying specific issues (e.g., "标题缺少搜索关键词，首屏信息密度太低，代码块无注释") → optimized version with before/after comparison table → re-publishing strategy.

Prompt 3: English-to-Juejin Adaptation#

User: "这篇英文技术博客写的不错，帮我改写成适合掘金发布的中文版 [paste or URL]" Expected Output: Localized article: not direct translation but culturally adapted — Chinese dev community references, relevant Chinese tech stack context, Juejin-appropriate humor/examples, SEO keywords in Chinese.

Prompt 4: Article Series Planning#

User: "我想写一个'微服务入门到实战'的系列，帮我规划前5篇的选题和结构" Expected Output: 5-article series plan: title per article, key takeaways, progression logic (foundation → practice → optimization → pitfalls → advanced), cross-linking strategy, estimated publishing cadence.

Prompt 5: Title A/B Test#

User: "这几个标题哪个在掘金最可能火？1) Go并发编程入门 2) 用了3年Go，才知道goroutine可以这样用 3) Go并发：从理论到实战的5个关键技巧" Expected Output: Per-title CTR potential analysis: pattern match, keyword power, emotion trigger, curiosity gap. Ranked recommendation with reasoning. "Title 2 has highest potential: curiosity gap + experience authority + practical value."

Prompt 6: Full Audit & Rewrite#

User: "把我最近的5篇文章都审计一遍，按优化优先级排序，然后从最差的那篇开始重写" Expected Output: Audit report ranking all 5 by optimization potential (estimated score gain) → full rewrite of the weakest article with all optimizations applied.

Real Task Examples#

Example 1: First-Time Juejin Author#

Scenario: Senior developer writing first Juejin article, unfamiliar with platform norms. Input: "写了篇微服务监控的文章，帮我看看在掘金能不能火 [paste draft: ~3000 chars, heavy code, light explanation]" Steps:

1. Diagnostic: Title 4/10 (generic), Code/Text 65% code (too much), Engagement 2/10 (no hooks), Readability 5/10.
2. Category benchmark: Backend category top articles average 45% code ratio, 4000-8000 chars.
3. Gap: need +25% text, engagement hooks, better title, visual suggestions.
4. Title: Generate 3 variants → user picks "微服务监控踩坑实录：这5个指标救了我3次".
5. Optimize: restructure with problem-first opening, add "why this matters" before each code block, add practical takeaways, insert 4 visual suggestion markers.
6. SEO: tags (微服务, 监控, Prometheus, 后端, 架构), primary keyword "微服务监控". Output: Optimized article (5200 chars, 42% code). Projected score improvement: 4.8/10 → 7.5/10. Publishing at Thursday 10:30.

Example 2: Company Tech Blog Pipeline#

Scenario: Company DevRel team needs to publish 4 articles/month on Juejin to build employer brand. Input: "这是我们团队本月要发的4篇草稿，帮我统一优化，风格要专业但不枯燥 [upload: 4 drafts]" Steps:

1. Per-article diagnostic scoring.
2. Cross-article consistency: ensure complementary topics, no overlap, coherent series if applicable.
3. Per-article optimization with shared tag strategy (build company tag presence).
4. Internal linking: cross-reference between articles where relevant.
5. Publishing calendar: stagger across 4 weeks with optimal days/times. Output: 4 optimized articles + publishing calendar + engagement plan + team author bio templates.

Example 3: Viral Article Reverse-Engineering#

Scenario: User wants to understand why a competitor's article went viral and apply lessons. Input: "这篇文章在掘金有5000赞，帮我分析它为什么火 [URL], 然后按同样的套路优化我的文章 [paste my draft]" Steps:

1. Analyze viral article: title pattern (numbered list + practical value), structure (problem → why it happens → solution → code → results), engagement hooks (3 discussion questions, 2 relatable scenarios), length (6200 chars, 38% code).
2. Extract replicable patterns → apply to user's draft.
3. Optimize: adopt same structure pattern, similar hook placement, comparable length. Output: "Viral pattern decoded: ① Hook with relatable pain point ② Show don't tell (benchmark data) ③ 'I wish I knew this earlier' framing. Applied to your article: [optimized version]. Estimated CTR uplift: +80%."

Scripts#

PathDescriptionscripts/optimizer.shMain CLI script — diagnose, title optimization, tags, publishing strategyschemas/input.schema.jsonJSON Schema for workflow input (title, task, tags)schemas/output.schema.jsonJSON Schema for workflow output (scores, variants, publishing)references/category_benchmarks.jsonJuejin category benchmarks for article patterns

CLI Usage#

bash

# Diagnose article title
bash scripts/optimizer.sh --title "Go并发编程入门" --task diagnose

# Generate optimized title variants
bash scripts/optimizer.sh --title "微服务监控实战" --task title

# Tag and SEO recommendations
bash scripts/optimizer.sh --title "Kubernetes教程" --task tags --tags "后端,Kubernetes"

# Publishing strategy
bash scripts/optimizer.sh --title "React性能优化指南" --task publish

# Full pipeline
bash scripts/optimizer.sh --title "Go并发编程入门" --task full --tags "后端,Go语言"

🚀 First-Success Path (3 Steps)#

1. Step 1: Run bash scripts/optimizer.sh --title "你的文章标题" --task diagnose
2. Step 2: Review the diagnostic scores and run bash scripts/optimizer.sh --title "你的文章标题" --task title for title variants
3. Step 3: Run bash scripts/optimizer.sh --title "你的文章标题" --task full --tags "标签" to receive the complete optimization report within 30 seconds

bash

# Quick start
bash scripts/optimizer.sh --title "Go并发编程入门" --task diagnose

Boundary Conditions#

ConditionBehaviorArticle <500 charactersFlag as too short for Juejin (min 1000 chars recommended); suggest expandingArticle >15,000 charactersWarn: "Very long — consider splitting into 2-part series for better engagement"Pure code dump (>80% code)Flag: "This reads like a gist, not an article. Add explanations and context."Non-technical contentWarn: "Juejin is a technical community. Entertainment/general content performs poorly."Article already publishedOffer "rescue" mode — analyze existing performance + suggest improvements for re-publish or next articleURL unreachable (Juejin blocks scrape)Request user to paste article text manuallyCategory has <10 reference articlesBroaden to adjacent categories for pattern analysisUser requests guaranteed viral resultRedirect: "Optimization improves probability, not guarantees. Even top authors have articles that underperform."

Error Handling#

Error CodeScenarioHandlingE-SCRAPE-BLOCKEDJuejin anti-scraping triggeredUse cached benchmark data; flag staleness; suggest manual category inputE-CATEGORY-UNKNOWNArticle topic doesn't match any Juejin categoryAsk user to specify closest category; use general tech benchmarkE-LOW-DATAInsufficient reference articles for pattern analysisBroaden time range or adjacent categories; flag lower confidenceE-ARTICLE-UNREADABLECan't parse article format (corrupted markdown, special chars)Offer plain text mode; warn formatting may be simplifiedE-TITLE-REJECTEDAll 3 title variants rejected by userGenerate 3 more with different patterns (try: controversy, question, how-to)E-OVER-OPTIMIZATIONToo many optimization suggestions overwhelm userOffer priority-ranked top-5 mode; defer others to "advanced optimization"

Security Requirements#

• Original content only: The optimizer enhances YOUR content. It does not generate content that plagiarizes existing Juejin articles. If pattern analysis detects high similarity to a specific published article, flag it.
• No automated posting: This tool generates optimized drafts and publishing advice. It does not post to Juejin on the user's behalf. User controls all publishing decisions.
• Article privacy: Draft content processed in-session only. Not stored or used for any other purpose.
• Juejin ToS compliance: All scraping is read-only and rate-limited. Do not circumvent Juejin's anti-scraping measures.
• Attribution integrity: If the user's article references external sources, ensure proper attribution is preserved and enhanced.
• Content safety: Reject optimization requests for content that violates Chinese internet regulations or Juejin community guidelines.