What does this skill do, and when should you use it?
Content Writer is an implementation-stage SEO/GEO skill in Aaron Marketing Skills with new and refresh modes. It drafts keyword-targeted articles, landing pages, and product copy, or analyzes an existing page for traffic or ranking decay and plans the update. The skill applies CORE-EEAT constraints, evidence boundaries, on-page SEO practices, and GEO guidance, but it does not calculate the CORE-EEAT score or issue the publishing veto. It fits teams that need a concrete SEO deliverable or refresh plan and can supply the relevant keywords, page context, URLs, or performance data.
It reads the content brief, target keywords, page intent, entity inputs, and Narrative and claims projections; refresh mode also reads the candidate URL or content, traffic and ranking history, publication dates, and competitor examples. New mode produces title options, a meta description, H1/H2/H3 structure, copy, FAQ content, a snippet-targetable block, and internal and external links, followed by SEO and CORE-EEAT self-checks. Refresh mode identifies decay drivers, prioritizes updates, drafts replacement content, sets a republishing strategy, and defines follow-up measurement. Metrics are labeled Measured, User-provided, Calculated, Estimated, or Proxy, and unsupported factual claims are marked [needs source].
- A content team needs a ready-to-use SEO article for a defined keyword, including titles, metadata, FAQs, and link structure.
- A marketer has an existing URL and publication date and wants to refresh an outdated article.
- A site owner sees declining traffic or rankings and needs a diagnosis plus prioritized update plan.
- A product team needs a search-targeted landing page or product description with evidence-safe claims.
How do you install this skill?
- Verify all statistics, competitor claims, traffic/ranking changes, and expected outcomes before publication; example targets are not validated forecasts.
- Confirm that the host provides the memory projections, registry-events.py, connectors, and CLAUDE_PLUGIN_ROOT referenced by the skill; otherwise stop or explicitly mark dependency_status instead of assuming availability.
- Resolve the mismatch between version 18.0.0 and SECURITY.md's stated current supported 17.x line.
- Chinese users may need localized templates, citation guidance, and output conventions; reachability and credentials for Baidu or other external services require separate validation.
- Shell / CLI
- Network access
- Local filesystem
Python 3
Install the repository in Claude Code with /plugin marketplace add aaron-he-zhu/aaron-marketing-skills, followed by /plugin install aaron-marketing@aaron. On other Agent Skills-compatible hosts, run npx skills add aaron-he-zhu/aaron-marketing-skills, or clone it with git clone https://github.com/aaron-he-zhu/aaron-marketing-skills. The source does not provide a command for installing this skill alone.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/aaron-he-zhu/aaron-marketing-skills.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/seo-geo/implement/content-writer" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
For new content, use a prompt such as Write an SEO-optimized article about [topic] targeting the keyword [keyword]. For a refresh, use Refresh this article for [current year]: [URL/content]. The argument format is [--mode new|refresh] <topic/keyword or URL of existing content>. An explicit mode is honored; otherwise, an existing URL plus a decline or staleness signal selects refresh, while a topic or keyword without a prior version selects new.
What are this skill's strengths and limitations?
- Supports both net-new SEO writing and existing-content refresh workflows.
- Produces concrete deliverables including titles, metadata, page structure, FAQs, snippet blocks, and links.
- Uses explicit evidence boundaries and metric labels, and prohibits fabricated facts, studies, dates, and attributions.
- Adds GEO-oriented definitions, quotable standalone statements, and Q&A structures.
- Works at Tier 1 with user-provided data and does not require paid tools or APIs.
- It does not calculate the CORE-EEAT score or replace content-quality-auditor as the publishing gate.
- It does not independently score AI-citation/GEO readiness or produce standalone meta/schema artifacts.
- Outcome analysis depends on supplied keywords, page information, performance data, and competitor examples; without analytics, decay findings are Estimated.
- The source does not show a dedicated test suite for this skill; the README only documents repository CI regression tests for the validator and shared contract.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| SEO Content Writer & Refresh Planner this page | 54 · Use with care | ★ 3k | 13d ago | Apache-2.0 |
| Blog Brand Context | 49 · Use with care | ★ 2.4k | 13d ago | MIT |
| Claude Blog Content Engine | 47 · Use with care | ★ 2.4k | 13d ago | MIT |
| Blog Brief Planner | 44 · Not recommended | ★ 2.4k | 13d ago | MIT |
| Blog Decay Analyzer | 44 · Not recommended | ★ 2.4k | 13d ago | MIT |
How did FollowSkills review this skill?
The skill treats pasted content, URLs, CSVs, and fetched content as untrusted, requires permission for memory writes, and gates indexing mutations behind dry-run plus explicit --live. However, it still involves external fetching, memory writes, registry proposals, and optional indexing notifications; data-flow, rollback, and sensitive-data controls are mostly delegated to repository-wide documents rather than this skill, and publisher identity is unverified, so points are deducted.
The two modes, input/output contract, stop conditions, missing-data handling, and blocked dependency state are well specified. Static review did not execute key paths, and the skill depends on memory projections, connectors, scripts, and host variables; the version also conflicts with SECURITY.md's stated supported line. The score therefore remains within the static ceiling and is reduced.
Triggers, new/refresh mode selection, argument hints, excluded neighboring skills, and behavior for missing URLs or keywords are clearly described. Chinese metadata and Baidu indexing are present, but the substantive instructions and templates are almost entirely English, with limited evidence for Chinese content workflows, host variation, and mainland-China network reachability, so points are deducted.
The skill has strong structure, quick starts, a contract, reference templates, checklists, license, and version metadata, with clear progressive disclosure. Maintenance ownership, update responsibility, and changelog path are not clear in the selected skill; repository version 18.0.0 conflicts with SECURITY.md's 17.x supported line, and several dependencies were not supplied as evidence, so points are deducted.
The skill defines directly usable deliverables for drafting and refreshing, including content, diagnosis, update plans, link recommendations, and handoff data. Static review cannot establish output correctness; results depend on user-supplied inputs, analytics, sources, and connectors, while example outcome targets lack verification. The score is therefore capped by static calibration and reduced.
The skill requires sourced, dated factual claims, evidence boundaries, and explicit metric labels, providing some auditability. However, the supplied files contain no reproducible key-path execution evidence or independent third-party validation for this skill; README claims about CI and verification are repository assertions, so only limited credit is warranted.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →