[geo-optimizer] GEO Audit Report — 2026-08-18

[github-actions[bot]]

GEO Audit Report — github/gh-aw

Audit Date: 2026-08-18
Run: https://github.com/github/gh-aw/actions/runs/32160801176

---

📊 Scores

Target	Score	Band
Docs site (github.github.com/gh-aw/)	44/100	Foundation
Docs sitemap average (20 pages sampled)	37.9/100	Foundation (16 foundation / 4 critical)
README (github.com/github/gh-aw)	55/100	Foundation

Note on robots.txt: the GEO scanner's robots_txt check reports found: false for the docs site because it probes the domain root (https://github.github.com/robots.txt), which doesn't include the /gh-aw/ GitHub Pages project-site base path. A direct, authoritative check of https://github.github.com/gh-aw/robots.txt confirms it does exist (HTTP 200) and explicitly allows all major AI crawlers (GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot, anthropic-ai, Google-Extended, etc.) plus a Sitemap: directive. This is a scanner limitation, not a real site gap — no action needed on robots.txt for the docs site.

---

✅ Top Strengths

• Meta tags: Docs site scores 14/14 — title, description, canonical URL, and full OpenGraph tags all present.
• Schema JSON-LD: Docs site includes WebSite, Organization, SoftwareApplication, and FAQPage types (13/16).
• Content structure: README scores 12/12 on content (word count, headings, links); docs site has a clear H1, 49 headings, and heading hierarchy.
• Signals: Docs site has lang="en", an RSS feed (/blog/rss.xml), and a freshness date — full 6/6.
• Brand & entity: Docs site links to Wikipedia, Wikidata, LinkedIn, and Crunchbase for Knowledge Graph disambiguation (7/10).
• Multimodal: Alt-text coverage is 100% and video captions are present (readiness: "good").
• CDN/bot access: No AI bots (GPTBot, PerplexityBot, Claude-SearchBot, Googlebot, etc.) are blocked or challenged at the CDN level.
• README: robots.txt (15/18) and llms.txt (14/18) checks both pass strongly at the repo level.

---

🚨 Critical Gaps

1. No llms.txt on the docs site (0/18) — no /llms.txt file exists to help AI systems index the site structure.
2. No AI Discovery signals on either the docs site or README (0/6 each) — missing /.well-known/ai.txt, /ai/summary.json, /ai/faq.json, /ai/service.json.
3. README has no Schema JSON-LD (0/16) — no WebSite/Organization/FAQPage structured data on the GitHub repo homepage.
4. Hidden text detected on both docs site and README (display:none/visibility:hidden with content) — flagged as a potential cloaking pattern that AI crawlers may penalize (contributes to the -3 negative penalty on both audits).
5. Keyword stuffing: "github" density is 2.7% (docs) / 3.6% (README) — vocabulary diversification recommended.
6. 4 of 20 sampled docs pages score "critical" (worst: /blog/9 at 33, /blog/11 at 35, /blog/8 at 35, /about at 35) — largely driven by weak schema/brand-entity scores on those pages.

---

🔧 Recommended Fixes

Ordered by estimated impact:

1. Create /llms.txt for the docs site (up to +18 pts) — run geo llms --base-url https://github.github.com/gh-aw to generate a structured AI-indexing file with H1, description, sections, and links.
2. Add Organization + WebSite JSON-LD schema to the README (up to +16 pts) — include name, url, logo, sameAs (Wikipedia/Wikidata/LinkedIn/Crunchbase), and FAQPage schema.
3. Add AI-discovery well-known files to the docs site and README (up to +6 pts each): /.well-known/ai.txt, /ai/summary.json, /ai/faq.json, /ai/service.json.
4. Remove or reveal hidden text (display:none/visibility:hidden elements with real content) on both the docs site and README to eliminate the cloaking-pattern penalty (-3 pts each).
5. Front-load key information in the first 30% of docs-site content for better AI snippet selection.
6. Add numerical data/concrete statistics to docs content (~+40% AI visibility per GEO heuristics).
7. Diversify vocabulary to reduce "github" keyword density below stuffing thresholds.
8. Add <link rel="canonical"> to the README page to prevent duplicate-content indexing issues.
9. Investigate the 4 critical-band docs pages (/blog/9, /blog/11, /blog/8, /about) for schema and brand-entity gaps specifically.

📋 Full Breakdown by Category

Docs site (https://github.github.com/gh-aw/) — score 44/100:

Category	Score	Max
Robots.txt	0*	18
llms.txt	0	18
Schema JSON-LD	13	16
Meta Tags	14	14
Content	7	12
Signals	6	6
AI Discovery	0	6
Brand & Entity	7	10
Negative penalty	-3	—

*See robots.txt note above — this is a scanner domain-root artifact; the actual /gh-aw/robots.txt is valid and AI-bot-friendly.

README (https://github.com/github/gh-aw) — score 55/100:

Category	Score	Max
Robots.txt	15	18
llms.txt	14	18
Schema JSON-LD	0	16
Meta Tags	11	14
Content	12	12
Signals	3	6
AI Discovery	0	6
Brand & Entity	3	10
Negative penalty	-3	—

Sitemap-wide averages (20 pages sampled of 224 discovered):

Category	Avg Score
Robots.txt	0.0
llms.txt	0.0
Schema JSON-LD	5.5
Meta Tags	14.0
Content	10.85
Signals	5.35
AI Discovery	0.0
Brand & Entity	3.1
Negative penalty	-0.9

Band distribution: 16 pages "foundation", 4 pages "critical" (out of 20 audited).

📄 Sitemap Page Scores

Top pages:

URL	Score	Band
/gh-aw (home)	44	foundation
/blog/2026-01-13-meet-the-workflows-continuous-improvement	44	foundation
/blog/2026-01-13-meet-the-workflows-continuous-refactoring	43	foundation
/blog/2026-01-12-welcome-to-pelis-agent-factory	41	foundation
/blog/2026-01-13-meet-the-workflows-advanced-analytics	41	foundation

Worst pages:

URL	Score	Band
/blog/9	33	critical
/blog/11	35	critical
/blog/8	35	critical
/about	35	critical
/blog/10	36	foundation

---

Automated audit powered by geo-optimizer-skill · Run logs

Generated by 🌍 GEO Optimizer Daily Audit · auto · 21.1 AIC · ⌖ 5.94 AIC · ⊞ 7.7K · ◷

• expires on Aug 21, 2026, 8:42 AM UTC-08:00