Agent skill
citation-radar
Audits how likely AI answer engines (ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews) are to cite your pages for the questions buyers ask, then ranks fixes by citation lift.
Filed under Content and SEO.
From varunk130/ai-gtm-skill-library · 37 skill entries · 6 · pushed 2026-09-19
What it does when it runs
Audits how likely AI answer engines (ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews) are to cite your pages for the questions buyers ask, then ranks fixes by citation lift. Use when: ai search visibility, generative engine optimization, geo audit, aeo, llm citations, are we cited in ai answers, ai overviews, answer engine optimization.
Automated analysis of the skill and the 0 files bundled beside it. A skill’s own description is written to be selected by an agent, so it describes the job and not the dependencies.
- Keys and connectors you must supply
- None found.
- Hosts it reaches
- No third-party host appears in the skill or its bundled files.
- Tool permissions it declares
- No
allowed-toolsin the frontmatter. It only issues instructions, so there is nothing to bound. - Actions present in the files
- None. Instructions only.
Install it
View source on GitHub ↗git clone --depth 1 --filter=blob:none --sparse https://github.com/varunk130/ai-gtm-skill-library.git /tmp/ai-gtm-skill-library git -C /tmp/ai-gtm-skill-library sparse-checkout set "gtm-skills/citation-radar" mkdir -p ~/.claude/skills/citation-radar cp -R "/tmp/ai-gtm-skill-library/gtm-skills/citation-radar/." ~/.claude/skills/citation-radar/
Picked up without a restart. A project skill of the same name is shadowed by your personal one. For one repository only, swap ~/.claude/skills for .claude/skills. Claude Code docs ↗
The folder is the same in every client that implements the format — 46 of them — so if yours is not above, only the destination changes.
The skill
Source on GitHub ↗Reproduced in full from varunk130/ai-gtm-skill-library/blob/158f12c03830dca454576eb1a78b1e4a34ae83e3/gtm-skills/citation-radar/SKILL.md, which is licensed MIT (repository). 482 words, 7 headings.
Citation Radar (CLEAR Framework)
Buyers now ask an assistant before they open a search results page. Classic SEO audits measure rank; this skill measures whether a page is the kind of source an answer engine can lift a sentence from and credit. It works from the buyer's questions backward, not from keywords forward.
Core Principle
Answer engines cite claims, not pages. A page earns citations when it contains short, specific, attributable statements that directly answer a question someone actually asks.
The CLEAR Framework
| Letter | Dimension | What It Checks | Weight |
|---|---|---|---|
| C | Claim density | Specific, checkable statements per 100 words (numbers, named entities, dates) | 0.25 |
| L | Liftability | Answer sits in the first 40-60 words of the section; lists and tables for comparisons | 0.15 |
| E | Evidence | Named author, dated update, primary sources, schema markup | 0.20 |
| A | Answer match | Does an H2/H3 or first sentence answer the buyer question verbatim? | 0.25 |
| R | Rarity | Information not found on the top competing pages (original data, pricing, benchmarks) | 0.15 |
Score each dimension 0-10. CLEAR score = Σ(weight × score) × 10 → 0-100.
- 75+ Citable - protect and refresh quarterly
- 50-74 Fixable - one or two dimensions drag it down
- <50 Invisible - rewrite around a specific question
Process
- Question set - collect 20-40 real buyer questions (sales calls, support tickets, community threads); tag each by funnel stage
- Map - match each question to the page that should answer it; unmatched questions are content gaps
- Probe - ask 2-3 assistants each question; record who is cited and which sentence they quote
- Score - apply CLEAR to every mapped page
- Prioritize - rank fixes by
question volume × (75 - current score), capped at 0 for pages already citable - Rewrite brief - for each top fix, name the question, the missing claim, and the proof to add
Output
Save to outputs/citation-radar-[domain]-[YYYY-MM-DD].md
| Artifact | Description |
|---|---|
| Question map | Buyer question → target page → current citer |
| CLEAR scorecard | Five dimension scores and total per page |
| Fix queue | Top 10 rewrites ranked by expected citation lift |
| Gap list | Questions with no page, ready for demand-engine |
Tips
- Probe, don't assume - the page an assistant cites today is the real competitor
- Dates matter - an undated page reads as stale to both people and models
- Tables win comparisons - "X vs Y" questions get answered from tables
Pairs With
- context-anchor - supplies Audience and Substantiation for rewrites
- demand-engine - turns the gap list into content
- competitive-battlecard - shows which competitor claims get cited instead of yours
Other skills for the same job
Different authors, same problem. Matched on the words in the skill name, across every library in the catalogue except this one.
- company-radar by Varnan-Tech · 660
- named-account-trigger-radar by zapier · 339
- trade-show-competitor-radar by LensmorOfficial · 48
- format-radar by Ajitesh-png · 0
Need help setting it up?
This page tells you what citation-radar does and what it needs. Cheetah builds the agent setup it runs inside: data, CRM, sequencing and the guardrails.
Book a call →The directory stays free. There is nothing gated behind this.