Free tools, open data, live dashboards.
One-page index of everything Garnet publishes for free. Tools you can use without signing up, datasets licensed CC0, dashboards updated daily by cron. Same infrastructure we run for paid clients — just open.
Free tools (use without signing up)
Paste a URL + buyer question. Gemini-with-Search returns who got cited instead. ~3 seconds.
On-prem Mac Mini MLX vs frontier API. Mini count, capex, monthly opex, break-even — math visible.
JSON-LD audit for any URL. Per-type score + paste-ready PR-shaped snippets with click-to-copy.
Live dashboards (re-rendered daily)
Every AI bot fetch on garnetgrid.com — OpenAI, Anthropic, Perplexity, Google, Apple, etc. Daily.
Same buyer query, 4 engines, 8 consulting categories, every day. Who each engine cites — raw.
Daily axe-core scan against WCAG 2.1 AA. 10 templates × 2 viewports. Per-rule + per-URL breakdown.
Every shipped change on garnetgrid.com, 120-day window. Categorized, filterable, git-driven.
Live 30-day citation arc on Garnet's own domain. Baseline 2026-05-10 → Day 30 2026-06-09.
Open registries
Machine-readable endpoints
Structured mirror of /llms.txt for agents. Lanes + prices + methodology URLs + live_metrics (real-time signal at request time).
Markdown reading-order for AI agents researching Garnet. Lane summaries + methodology + onboarding pointers.
50 most-recent publications: insights, case studies, methodology, vs-alternatives, onboarding, live dashboards. Daily.
All ~91 indexable pages with lastmod, changefreq, priority.
Raw CC0 datasets (no auth, no rate limit)
30-day rolling AI-bot fetch aggregate. By vendor, bot, purpose, page path. Daily refresh.
8 categories × 4 engines × 30-day rolling. Per-engine top-cited domains + cross-engine consensus.
Latest axe-core scan: WCAG 2.1 AA strict + moderate counts, per-rule + per-URL breakdown.
120-day commit history grouped by date, classified by category prefix.
Live in-house GEO measurement state — baseline, days_elapsed, phase, per-engine breakdown.
All 8 paste-ready JSON-LD packs as structured data (id, priority, after-block).
Why publish all of this. Garnet sells citation engineering, architecture audit, Discord ops, and on-prem MLX inference. If we can't measure and publish those things on our own domain first, we shouldn't be selling them. The infrastructure on this page is the same infrastructure paid subscribers get on their own customer Cloudflare accounts.
What we don't publish here. Per-customer data, raw webhook payloads, customer R2 contents, individual request logs. Garnet stores diffs in a control plane; customer-owned data stays in customer-owned R2 buckets. Sovereignty by default.