aidenlx/zotlit / .agents/skills/cloudflare-web-analytics
cloudflare-web-analytics
VerifiedQuery Cloudflare Web Analytics RUM and Core Web Vitals through the GraphQL Analytics API. Use when a Web Analytics dashboard URL or site tag is supplied, or when diagnosing CLS, LCP, or INP by page, element, browser, device, country, or host.
Query Cloudflare Web Analytics
Use field data to choose the page and environment to reproduce. The GraphQL schema is the source of truth: the RUM datasets are beta and can change independently of static guidance.
1. Resolve the query scope
Extract these values from the request or dashboard URL:
accountTag: the 32-character account ID in the dashboard path.siteTag: the value ofsiteTag~in.- Time range: preserve explicit dashboard filters. Otherwise choose a bounded recent range and state it in the result.
- Bot filter: map
excludeBots=Yestobot: 0.
Query requestHost before narrowing to a deployment. One site tag can contain more than one
host, including production and pre-release hosts.
query Hosts($accountTag: string!, $siteTag: string!, $start: Date!, $end: Date!) {
viewer {
accounts(filter: { accountTag: $accountTag }) {
rumWebVitalsEventsAdaptiveGroups(
filter: {
siteTag: $siteTag
date_geq: $start
date_leq: $end
bot: 0
}
limit: 50
orderBy: [count_DESC]
) {
count
avg {
sampleInterval
}
dimensions {
requestHost
}
}
}
}
}
This step is complete when the account, site tag, ISO date range, bot policy, and target host are explicit.
2. Query the current schema
Send read-only GraphQL queries with mcp__cloudflare_api__execute:
async () => {
const response = await cloudflare.request({
method: "POST",
path: "/graphql",
body: { query, variables },
});
return response;
}
The account-level RUM datasets currently include:
rumWebVitalsEventsAdaptiveGroups— aggregated Core Web Vitals.rumWebVitalsEventsAdaptive— event-level Core Web Vitals.rumPageloadEventsAdaptiveGroups— aggregated page-load metrics.rumPerformanceEventsAdaptiveGroups— aggregated performance events.
Use GraphQL introspection before relying on a dataset, field, filter, or order value that has not already succeeded in the current run. On a schema error, read schema-discovery.md and rebuild the query from the returned types.
3. Establish the CLS baseline
Keep the first grouping small. Every selected dimension partitions the result further.
query ClsBaseline(
$accountTag: string!
$siteTag: string!
$host: string!
$start: Date!
$end: Date!
) {
viewer {
accounts(filter: { accountTag: $accountTag }) {
overall: rumWebVitalsEventsAdaptiveGroups(
filter: {
siteTag: $siteTag
requestHost: $host
date_geq: $start
date_leq: $end
bot: 0
cumulativeLayoutShift_geq: 0
}
limit: 1
) {
count
avg {
cumulativeLayoutShift
sampleInterval
}
quantiles {
cumulativeLayoutShiftP50
cumulativeLayoutShiftP75
cumulativeLayoutShiftP90
cumulativeLayoutShiftP95
}
}
culprits: rumWebVitalsEventsAdaptiveGroups(
filter: {
siteTag: $siteTag
requestHost: $host
date_geq: $start
date_leq: $end
bot: 0
cumulativeLayoutShift_gt: 0
}
limit: 25
orderBy: [quantiles_cumulativeLayoutShiftP75_DESC]
) {
count
avg {
cumulativeLayoutShift
sampleInterval
}
quantiles {
cumulativeLayoutShiftP50
cumulativeLayoutShiftP75
cumulativeLayoutShiftP90
cumulativeLayoutShiftP95
}
dimensions {
cumulativeLayoutShiftElement
cumulativeLayoutShiftPath
requestPath
}
}
}
}
}
Pass variables separately:
{
"accountTag": "ACCOUNT_ID",
"siteTag": "SITE_TAG",
"host": "www.example.com",
"start": "2026-08-01",
"end": "2026-08-27"
}
For a drill-down, add only the dimension that tests the current question, such as
deviceType, userAgentBrowser, or countryName. Filter to one requestPath or
cumulativeLayoutShiftElement before adding several dimensions.
4. Interpret sampled field data
- Use CLS p75 for the site-level verdict. Keep p50, p90, and p95 visible to show the distribution.
- Report
avg.sampleIntervalwith counts. A value greater than one means the query used sampled data, so a returned count is not a raw unsampled beacon count. - Treat
cumulativeLayoutShiftElementas the element that moved. An ancestor, font swap, injected sibling, or page-wide reflow can be the cause. - Treat equal extreme values across unrelated selectors as evidence of one shared reflow and test shared layout, fonts, theme initialization, and hydrated chrome first.
- Separate observations from inferences. A selector in RUM is evidence that it shifted, not proof that its component caused the shift.
5. Reproduce the selected segment
Open a high-impact path in Chrome at the matching device class and browser conditions. Record
a cold-load performance trace, then inspect the CLSCulprits insight. Use the trace to find
the source element, movement time, and preceding DOM or style change.
When source is available, map the trace back to the shared layout before changing individual reported elements. Re-run the same trace after the change and compare the CLS value and shift events.
This step is complete when one reproducible browser event explains the field-data pattern and the post-change trace removes or materially reduces that event.
Report
Include:
- Query scope: site tag, host, date range, and bot filter.
- CLS p75 and sample interval.
- Top paths and shifted selectors with their counts and distribution.
- The browser-trace cause, clearly marked as reproduced evidence.
- The changed files and the post-change measurement.
Use Cloudflare's current documentation for product behavior and the live GraphQL schema for API names. The Web Analytics FAQ documents GraphQL sampling, and the Core Web Vitals guide defines the dashboard's metric and debug views.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| deprecation-and-migrationaddyosmani/agent-skillsManages deprecation and migration. Use when removing old systems, APIs, or features. Use when migrating users from one implementation to another. Use when migrating a database schema in production, such as renaming or dropping a column without downtime (expand/contract). Use when deciding whether to maintain or sunset existing code. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| performance-optimizationaddyosmani/agent-skillsOptimizes application performance across frontend, backend, queries, and databases. Use when performance requirements exist, when you suspect performance regressions, when Core Web Vitals or load times need improvement, when N+1 query patterns need fixing, or when profiling reveals bottlenecks. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| Agent Cost Reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json. | Stars: 117k | Updated | Claude CodeCodex | Verified | Install |
| Agent Cost Reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json. | Stars: 117k | Updated | Claude CodeCodex | Verified | Install |