A durable history of this account's own cache performance.
This repo has a pattern: take something real and durable, and turn it into a small D1-backed primitive the rest of the ecosystem can reference -- subagentcontracts.com did it for the Contract, subagenttasks.com did it for the Task, seeded with this session's own 44 real tasks. subagentcache.com does the same thing for Cloudflare cache/CDN analytics: a growing, queryable history of real numbers pulled from this exact Cloudflare account (e6294e3ea89f8207af387d459824aaae), not a synthetic dashboard mockup.
Latest snapshot manual dashboard paste · captured 2026-07-01 18:53:13 · detail →
Today's real numbers, honestly labeled
The row above is not a live API call -- it's manual dashboard paste: the account owner opened their own Cloudflare Analytics dashboard, read the Cache widget, and pasted the exact figures in on 2026-07-01. That's a deliberate, load-bearing design choice, not a shortcut: every snapshot this site ever stores carries a source column that is either manual_dashboard_paste or graphql_api, and every page that renders a snapshot renders that badge alongside it. A manually-typed number is never presented as if a machine verified it, and vice versa.
Live ingestion is a documented next step, not done yet
The schema is designed today so a future scheduled Worker can write graphql_api-sourced rows into this exact same table, calling Cloudflare's own GraphQL Analytics API (httpRequests1dGroups dataset, sum{requests, cachedRequests, bytes, cachedBytes}) with a scoped CF_API_TOKEN secret (Zone Analytics Read permission -- every zone on this account, including this one, already carries #analytics:read). That cron path is not wired up in this build -- see the README for the exact next steps. Until then, new snapshots arrive the same way today's did: a real paste, honestly labeled.
Explore
Browse the full snapshot history → · Read why the provenance column is the whole point → · Hit the JSON API →