/seo-content-brief

name: seo-content-brief
description: Generate a writer-ready SEO content brief from a target domain and topic. Pulls domain overview, competitors, keyword gaps, SERP analysis, related and question keywords, AI Search citations, and existing internal-link sources, then synthesises a complete editorial brief a freelance writer can start from immediately. Use when the user asks for a content brief, blog brief, article outline, editor brief, or wants to capture organic traffic their competitors have.

> Example output: [examples/seo-content-brief-vercel-rate-limiting-20260514/BRIEF.md](../../examples/seo-content-brief-vercel-rate-limiting-20260514/BRIEF.md)

Content Brief

Turn a domain plus a topic intent into a complete content editor brief: target keyword, title options, H2/H3 structure, content gaps the current top results miss, internal linking plan, AI search angle, and estimated traffic potential.

Prerequisites

• SE Ranking MCP server connected.
• Claude's `WebFetch` tool available (used for top-3 content teardown).
• User has provided: (a) target domain, (b) market/country (default: `us`), and optionally (c) a seed topic or intent. If no seed topic is given, discover the best opportunity from the keyword-gap step.

Process

1. **Domain overview** `DATA_getDomainOverviewWorldwide`, `DATA_getDomainKeywords`

• Pull organic traffic, top countries, and top 100 organic keywords in the target market.
• Save the raw JSON and a human summary.

2. **Competitor discovery** `DATA_getDomainCompetitors`

• Identify the top 5 organic competitors by shared keywords in the target market.
• Save a one-line positioning note per competitor.
• **Note:** the upstream API does not support `limit`/`offset`, so this call returns the full set (~60KB for popular domains) and the MCP harness writes it to a file. Read that file path, parse the `{data: [...]}` JSON, sort by `common_keywords` desc, and take the top 5.

3. **Keyword gap analysis** `DATA_getDomainKeywordsComparison`

• Pull keywords the competitors rank for that the target domain does not.
• Filter: informational intent, search volume > 1,000/mo, keyword difficulty < 40.
• Save the filtered gap list sorted by volume.

4. **Pick the best topic cluster**

• From the gaps, select one topic. Justify the pick with: traffic potential, difficulty, relevance to the target domain's product. Surface reasoning to the user before proceeding.

5. **SERP and keyword deep-dive for the chosen topic**

• `DATA_getSerpResults` for the top 10 organic + SERP features (AIO, PAA, Featured Snippet, Video).
• `DATA_getRelatedKeywords` and `DATA_getSimilarKeywords` for expansion.
• `DATA_getKeywordQuestions` for People-Also-Ask and question-based variations.
• `DATA_getAiOverview` + `DATA_getAiOverviewLeaderboard` to see which brands LLMs cite today for the topic.

6. **Top 3 content analysis** `WebFetch` (always) + `mcp__firecrawl-mcp__firecrawl_scrape` (when available)

• **WebFetch first** (free, instant): pull markdown for the top 3 ranking URLs. Extract H1/H2/H3 spine, word count per article, shared subtopics, gaps, and prose-level formatting patterns.
• **Firecrawl second** (3 Firecrawl credits — 1 per top-3 winner) — recovers what WebFetch's markdown can't show:
• From `metadata`: `<title>` length (the real string, not markdown's first heading), meta description length, `og:title`, `og:description`, `og:image`, `twitter:card`.
• From the returned `html`: every `<script type="application/ld+json">` block. Parse and list `@type`s per winner (Article, FAQPage, BreadcrumbList, Product, etc.) — these become the "schema baseline" for the new article.
• On-page signals: hero-image presence, byline structure (`<a rel="author">`, `<meta name="author">`), table count, code-block count.
• **Classify the brief's template** against the 8-template map in `references/intent-template-map.md`. Cross-reference: (a) the dominant page type across the SERP top-10 (use the heuristics in `skills/seo-sxo/references/page-type-patterns.md`), (b) PAA patterns from step 5, (c) the keyword's intent classification from `DATA_getRelatedKeywords`. Pick one of: `ultimate-guide` / `how-to` / `listicle` / `explainer` / `comparison` / `review` / `best-of` / `landing-page`. If the SERP is split across types, follow the MIXED rule (see Tips). The chosen template determines the recommended H1/H2 outline shape and word-count floor — record both the template and a one-sentence justification in `BRIEF.md`.
• **If Firecrawl unavailable (or `--no-firecrawl` passed):** WebFetch portion runs unchanged. The brief's "Top 3 winners — on-page benchmark" subsection (see Output) emits `(skipped — Firecrawl required for schema/og:* on competitor pages)`. Template classification still runs from WebFetch + SERP data.

7. **Internal linking plan**

• `DATA_getDomainKeywords` filtered to the target domain plus WebFetch of 5 high-ranking pages on topically adjacent queries.
• For each: propose an anchor text and the section of the new post it belongs in.

8. **Synthesise the brief** (see Output Format).

Output format

Create a folder `seo-content-brief-{target-slug}-{YYYYMMDD}/` with the synthesised brief at the top level and step files in `evidence/`:

seo-content-brief-{target-slug}-{YYYYMMDD}/
├── BRIEF.md                        (writer-ready synthesis — primary deliverable; inlines 01-domain-overview, 02-competitors, 06-internal-links into a "Context" section)
└── evidence/
    ├── 01-domain-overview.md       (DATA_getDomainOverviewWorldwide raw — preserved for reproducibility)
    ├── 02-competitors.md           (DATA_getDomainCompetitors raw)
    ├── 03-keyword-gaps.md          (DATA_getDomainKeywordsComparison filtered)
    ├── 04-serp-and-keywords.md     (DATA_getSerpResults + related/question keywords)
    ├── 05-content-analysis.md      (top-3 winners' H-spine + on-page benchmark)
    └── 06-internal-links.md        (target