What does this skill do, and when should you use it?
This skill is the SEO/GEO survey-stage keyword research component of Aaron Marketing Skills. It supports research for a new page, topic, or campaign by organizing keywords around search volume, difficulty, intent, and topic clusters. It separates Quick Win, Growth, and GEO opportunities and requires metrics to be labeled by evidence type. The deliverable includes a prioritized keyword brief, topic clusters, content calendar, and handoff summary.
It reads a topic or seed keyword, target market and language, business goal, site DR, and user-provided or tool-derived metrics. It runs eight announced phases: Scope, Discover, Variations, Classify, Score, GEO-Check, Cluster, and Deliver. Optional data paths include SEO tools, Search Console, a Python Google Autocomplete helper, keyless Firecrawl SERP sampling, and a Wikipedia pageview demand proxy. It produces a prioritized keyword brief with volume, difficulty, and intent or labeled N/A; pillar-and-cluster hubs; Quick Win, Growth, and GEO opportunities; a content calendar; and a reusable research handoff.
- An SEO practitioner researching seed, long-tail, and commercial-intent terms for a new product page.
- A content team structuring a pillar page and supporting cluster topics for a new subject area.
- A marketing lead conducting initial demand research from pasted data or free sources without a paid SEO platform.
- A GEO team identifying question, definition, comparison, list, and how-to queries suited to AI answers.
- A site team mining Search Console rankings around positions 5–20 for striking-distance opportunities.
How do you install this skill?
- Before execution, confirm authorization to write memory and save results, and review target paths and sensitive data.
- When volume or difficulty is unavailable, retain N/A or clearly labeled estimates; do not treat Wikipedia attention, SERP observations, or sample numbers as search volume.
- Separately confirm reachability, terms, and data-egress implications for external services such as Firecrawl and Google Autocomplete; this review executed no scripts or connectors.
- Shell / CLI
- Network access
- Local filesystem
Python 3
For an Agent Skills-compatible host, run: npx skills add aaron-he-zhu/aaron-marketing-skills -s keyword-research. You can also clone the full repository with: git clone https://github.com/aaron-he-zhu/aaron-marketing-skills. In Claude Code, add the marketplace with /plugin marketplace add aaron-he-zhu/aaron-marketing-skills, then install with /plugin install aaron-marketing@aaron.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/aaron-he-zhu/aaron-marketing-skills.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/seo-geo/survey/keyword-research" ~/.claude/skills/
rm -rf "$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?
Use a prompt such as: Research keywords for my SaaS product targeting small teams, or Research keywords for [topic/product/service]. Provide a topic or seed keyword and, where available, the market/language, audience, business goal, site DR, and known metrics. The skill announces each [Phase X/8: Name] and returns the keyword brief, opportunity groups, topic clusters, content calendar, and next steps.
What are this skill's strengths and limitations?
- A fixed eight-phase workflow covers discovery, intent classification, scoring, GEO checks, and clustering.
- It explicitly labels metrics as Measured, User-provided, or Estimated and uses N/A when required data is unavailable.
- It supports keyless workflows with pasted data and provides Python autocomplete, SERP-sampling, and topic-demand helpers.
- It defines a handoff to competitor-analysis and requires at least three prioritized Quick Win, Growth, or GEO opportunities.
- Without an SEO tool or Search Console, search volume and difficulty may remain N/A rather than being invented.
- The autocomplete helper uses an unofficial Google Autocomplete endpoint; the keyless Firecrawl path is described as having about 1,000 credits per month.
- The skill focuses on keyword research and content planning; competitor-relative content gaps, content writing, and technical SEO belong to other skills.
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 |
|---|---|---|---|---|
| Keyword Research & Content Planning this page | 54 · Use with care | ★ 3k | 13d ago | Apache-2.0 |
| Blog Cluster Engine | 53 · Use with care | ★ 2.4k | 13d ago | MIT |
| Competitor SEO & GEO Analyst | 51 · Use with care | ★ 3k | 13d ago | Apache-2.0 |
| SEO Content Brief Generator | 52 · Use with care | ★ 19k | 22d ago | MIT |
| Blog Calendar Editorial Planner | 53 · Use with care | ★ 2.4k | 13d ago | MIT |
How did FollowSkills review this skill?
The skill distinguishes measured, user-provided, and estimated data, keeps integrations optional, and repository security material describes read-only defaults and non-persistent keys. However, it writes deliverables and memory without explicit per-write confirmation, least-write scoping, rollback, or sensitive-data handling; deductions apply. The publisher is unverified, so identity remains unknown.
The eight-phase workflow, N/A rules, and metric labels support consistent use. However, the skill depends on external references, scripts, and optional connectors, with no skill-specific key-path tests, abnormal-input handling, or diagnosable failure feedback; static calibration keeps this at or below 10.
Triggers, inputs, outputs, and the non-fit boundary for competitor coverage gaps are reasonably clear, including Chinese trigger terms. Still, market, language, and data-source fit depend on user data or external services; degradation without tools is incompletely specified, and mainland-China reachability of some overseas services is not established.
The skill includes structured metadata, an Apache-2.0 license, versioning, quick starts, a contract, references, an example, and next-skill guidance. It does not clearly provide a changelog, maintenance owner, stable-parameter policy, or troubleshooting path within the assessed skill, and relies on repository-level documentation, so full marks are not justified.
The workflow covers discovery, intent classification, scoring, clustering, and delivery, requiring each candidate to include metrics or N/A, which could yield a usable research brief. However, the example and process are not supported by real execution evidence, and missing volume or difficulty data leaves core outputs dependent on external input; static calibration caps this at 7.
The formula, intent taxonomy, source labels, and N/A constraints provide limited auditability. The example is documentation-only, while repository tests mainly cover architecture and a generic validator rather than this skill's key paths or independent third-party results; static calibration caps this at 5.
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 →