GEO Platform Optimizer
Audits and scores a website against five AI search platforms — Google AI Overviews, ChatGPT, Perplexity, Gemini, and Bing Copilot — with per-platform checklists and a consolidated optimization report.
The skill mainly produces checklists and scoring rubrics; declared tools are Read/Grep/Glob/Bash/WebFetch/Write, with no sensitive-data collection or destructive operations, so the risk surface is small. Deducted for: no disclosure of external fetch data flow, no user-confirmation mechanism or rollback notes, and Bash/Write permissions not scoped to least privilege.
The five platform checklists and 0-100 rubrics are structurally consistent and self-coherent with a defined output format. Deducted for: no test coverage of this skill itself (repo tests only cover fetch_page.py SSR heuristics), many rubric items rely on subjective human judgment, and no defined behavior or failure feedback on abnormal input; per calibration, reliability above 10 is unreachable in a static review.
Scenario is clear (platform-specific AI search audits across five platforms) with reasonable trigger semantics and stated audiences. Deducted for: no declared capability boundaries or non-fit ranges, no Chinese-language support, and core function depends on Google/ChatGPT/Bing Copilot platforms that are not directly reachable from mainland China networks, which matters for FollowSkills users.
Has version 1.0.0, MIT license, structured frontmatter, and a well-layered flow (checklists → rubrics → summary → output template). Deducted for: author field is only 'geo-seo-claude', no changelog, no explicit maintenance ownership or update path, and README mixes in Skool community marketing plus uncited market statistics, leaving hidden assumptions.
As a static checklist it can guide a structured audit and emit a fixed-format GEO-PLATFORM-OPTIMIZATION.md report, offering some marginal value. Deducted for: many scoring items depend on unverifiable subjective judgment, headline claims lack sources, and no evidence that outputs are directly usable without significant rework; static-only confidence caps this at a conservative score.
All key statistics (92% AIO citations from top-10, Wikipedia 47.9%, Reddit 46.7%, 11% overlap) are unattributed and untraceable. No repo tests cover this skill path, no third-party execution evidence, no cross-source corroboration, and fact/inference are not separated, so only a low score is supportable.
- Static review only; no checklist flow was executed. Scores reflect source files at the pinned revision with low confidence.
- All platform citation statistics (92%, 47.9%, 46.7%, 11%, etc.) are unattributed and should be independently verified before reliance.
- Many rubric items require subjective human judgment; different assessors may produce widely divergent scores.
- The skill depends on Google, ChatGPT, and Bing Copilot, which are not directly reachable from mainland China; Chinese users need their own network arrangements.
- No Chinese-language support or guidance for auditing Chinese-language content.
- The README's Skool community promotion and market statistics are marketing material, not evidence of skill quality, and should not drive adoption decisions.
What does this skill do, and when should you use it?
This is one of the bundled sub-skills in the geo-seo-claude repository, located at skills/geo-platform-optimizer/SKILL.md, version 1.0.0. Its premise: only about 11% of domains are cited by both ChatGPT and Google AI Overviews for the same query, because each platform uses different indexes and source preferences. The skill provides, per platform, an explanation of how sources are selected, a ten-item optimization checklist, and a 0-100 scoring rubric, then synthesizes everything into a GEO-PLATFORM-OPTIMIZATION.md report with scores, gaps, and a phased action plan. It is fundamentally a structured audit playbook that Claude executes step by step against a target website.
Given a target URL and its industry, it audits the site platform by platform: Google AIO (question-based headings, direct first-paragraph answers, tables/lists, FAQ sections, cited statistics, dates and author bylines, page depth); ChatGPT (Wikipedia/Wikidata entity presence, Bing index coverage, Reddit and YouTube footprint, entity consistency, 2000+ word comprehensive content); Perplexity (Reddit activity, forum mentions, content freshness, original research, quotable standalone paragraphs); Gemini (Knowledge Panel, Google Business Profile, YouTube chapters, Schema.org markup, image optimization, E-E-A-T signals); Bing Copilot (Bing Webmaster Tools, IndexNow, LinkedIn/GitHub presence, meta descriptions, exact-match keywords, sub-2-second page load). Each platform receives a 0-100 score from its rubric, and the skill writes a Markdown report with a combined GEO score, status tiers (Strong 70+, Moderate 40-69, Weak 0-39), and quick-win/medium-term/strategic action items.
- A GEO/SEO consultant preparing an AI search visibility diagnosis who needs per-platform scores and a client-ready deliverable
- A content team whose site ranks well on Google but gets no ChatGPT citations and wants to pinpoint cross-platform gaps
- A SaaS company assessing its entity coverage across Wikipedia, Wikidata, Reddit, and YouTube
- A local business owner wanting to know which Google/Microsoft ecosystem assets are missing for Gemini and Bing Copilot citations
- A marketing team building a quarterly plan who needs actions tiered into this-week quick wins, this-month, and this-quarter
What are this skill's strengths and limitations?
- Covers five major AI search platforms and spells out how their selection logic differs (e.g., ChatGPT builds on Bing's index, Perplexity leans heavily on Reddit)
- Each platform gets an actionable ten-item checklist plus a quantified rubric, making outputs reproducible and client-deliverable
- Distinguishes universal optimization actions from platform-specific priorities, avoiding one-size-fits-all advice
- Fixed report format with status thresholds and a phased action plan, easy to track against over time
- Key statistics in the checklists (e.g., Wikipedia at 47.9% of ChatGPT citations, Reddit at 46.7% for Perplexity) are unsourced and unverifiable
- Audit quality depends on Claude's live inspection and judgment; scores include subjective elements and the repo ships no test suite
- Several rubric items (original research, community activity, Knowledge Panel) cannot be assessed automatically from code alone; execution quality depends on the operator
- Frontmatter uses Claude Code's allowed-tools declaration, which needs manual edits to port elsewhere
- The repo description pushes a paid Skool community; evaluate independently before adopting the commercial angle
How do you install this skill?
The skill ships inside the geo-seo-claude collection. Install the whole repo: on macOS/Linux run curl -fsSL https://raw.githubusercontent.com/zubair-trabzada/geo-seo-claude/main/install.sh | bash, or git clone and run ./install.sh; on Windows use install-win.sh from Git Bash. Requires Python 3.8+, Claude Code CLI, and Git; Python dependencies install into an isolated venv at ~/.claude/skills/geo/.venv/. The repo does not document installing a single sub-skill out of the 16 on its own.
How do you use this skill?
Inside Claude Code, invoke the collection's /geo platforms <url> command to trigger platform-specific optimization. Alternatively, give Claude a target URL and industry and ask it to follow this skill's flow: collect URL and topic → run each platform checklist → score each on the 0-100 rubric → generate GEO-PLATFORM-OPTIMIZATION.md. SKILL.md defines no trigger beyond the collection's slash commands and does not specify how platform data points (Bing indexing, Reddit mentions) are actually fetched.
How does this skill compare with similar options?
It is one of 16 sub-skills in the geo-seo-claude collection, complementary to but distinct from siblings like geo-citability (AI citation readiness scoring), geo-brand-mentions (brand mention scanning), and geo-schema (structured data); the README names no external competitors.