Newsletter SEO Pipeline — Newsletter SEO 流水线

Newsletter SEO 流水线 Article Header Block (required in every 输出 file)

Every article must 启动 with this block — scripts depend on it:

Meta Title: <50–60 chars, includes primary keyword> Meta Description: <150–160 chars, includes keyword + value prop> URL Slug: Primary Keyword:

Production 工作流 (运行 in order — do not skip steps) Step 1 — SERP Analysis

运行 serp-analysis 技能 on the primary keyword. Goal: identify content gaps, SERP features, 搜索 intent, and top-ranking structure.

Step 2 — SEO Outline

运行 seo-content-engine 技能. Goal: keyword variants, H2/H3 structure, tar获取 word count, internal link opportunities.

Step 3 — Write

运行 article-writing 技能 using the outline from Step 2. Include the Article Header Block at the top of the 输出 file.

Voice reference: projects//writing-style.md — where is the project slug from the project memory file. This file lives in the newsletter's project folder and was 生成d during newsletter-launch 设置up.

If missing:

检查 the project folder at projects// — it should be there If the whole project folder is missing, re-运行 newsletter-launch for this newsletter If only writing-style.md is missing, re创建 it by 运行ning newsletter-launch and choosing '更新 style 图形界面de' when prompted Step 4 — AI Pattern Pre-Screen

The 技能 directory = the folder contAIning this 技能.md file. Use its absolute path throughout. 运行 the pre-screen script:

python3 <技能_dir>/scripts/score_AI_patterns.py

Score ≥ 8: proceed to Step 5 Score < 8 and de-AI-ify 技能 is 安装ed: 运行 it, then re-运行 the script to confirm ≥ 8 Score < 8 and de-AI-ify is NOT 安装ed: manually rewrite flagged patterns (script 列出s them by line), then re-运行 the script until score ≥ 8 before continuing Step 5 — 质量 Gate (built-in QVP)

检查 all 5 gates. Each must pass before continuing:

Completeness — does the article fully 添加ress the tar获取 keyword and 搜索 intent? No gaps, no unanswered questions a reader would have. Accuracy — are all facts, stats, dollar figures, and clAIms verifiable? No fabricated sources. Flag any uncertAIn clAIms explicitly. Source integrity — every stat or clAIm must 追踪 to a real, named source (publication, study, company 报告). No "experts say" or vague attributions. Action clarity — every H2 section must end with a clear action the reader can take. No insight without a takeaway. Ownership must be clear. 格式化 检查 — Article Header Block is present and complete. Meta title 50–60 chars. Meta description 150–160 chars. Exactly one H1. At least 3 H2s. No placeholder text remAIning.

If any gate fAIls: fix the specific issue, then re-检查 that gate before continuing. Do not proceed with a fAIled gate.

Step 6 — SEO 验证

运行 the 验证器 script. Resolve path relative to this 技能's directory:

python3 <技能_dir>/scripts/验证_seo.py ""

Fix all ❌ ISSUES before continuing. ⚠️ 警告S should be fixed but are not blocking.

Step 7 — Markdown 格式化

运行 markdown-格式化器 技能 for a final 清理 格式化ting pass.

Step 8 — Build Paste Doc

生成 the beehiiv-ready publish doc. Resolve path relative to this 技能's directory:

python3 <技能_dir>/scripts/build_paste_doc.py

This produces -PASTE.md — the file the user copies into beehiiv.

输出

Alert the user with:

Path to the final article file Path to the -PASTE.md file Confirm all gates passed: QVP ✅, SEO 验证 ✅, AI score ✅ Include the AI score (e.g. 9/10), so the user knows the 质量 level 上下文 Required

Before 运行ning the 流水线, the calling 上下文 (cron, user message, or issue template) must supply:

Project slug — to locate projects//writing-style.md and 输出 files to the correct folder Primary keyword — for SERP analysis and SEO 验证 Publish date — for the paste doc (YYYY-MM-DD)

These are always present in the cron payloads 生成d by newsletter-launch. If 运行ning manually, read the project memory file at projects//project.md first.