agentleFS
catesandrewcatesandrew/skills/skills/chrome-audit-network/SKILL.md

chrome-audit-network

Skill1 starsChanged 5 days agoMITView on GitHub

Use when you need to capture and summarize network activity for a page or user flow via Chrome DevTools MCP — identifies slow requests, large payloads, and API call patterns.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If filePath is omitted, auto-name: ./tmp/network-<sanitized-host>.png

    line 22
  2. Very specific:

    Wait ${input:duration} ms (or until network appears settled).

    line 39

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

chrome-audit-network

Navigate to a URL, collect Performance Resource Timing entries, filter/summarize, and optionally screenshot. Best for diagnosing slow page loads, unexpected API calls, or payload size issues.

Requires: Chrome DevTools MCP server running and connected.

Inputs

VariableDescriptionExample
${input:url}Page URL to audithttp://localhost:4200, https://app.example.com/dashboard
${input:filter}Regex filter for resource names (optional).*(api|graphql).*, .*\.(js|css)$
${input:duration}Wait duration in ms (optional, default: 5000)3000
${input:filePath}Screenshot save path (optional)/tmp/network-dashboard.png

If filePath is omitted, auto-name: ./tmp/network-<sanitized-host>.png

sanitize(x): lowercase; keep a–z 0–9 - _; replace other chars with -.

Procedure

1. Navigate

mcp__chrome__navigate_page: { "url": "${input:url}" }

2. Clear existing timings

performance.clearResourceTimings();

3. Wait for activity

Wait ${input:duration} ms (or until network appears settled).

4. Collect resource entries

(() => {
  return performance.getEntriesByType('resource').map(e => ({
    name: e.name,
    initiatorType: e.initiatorType,
    transferSize: e.transferSize,
    encodedBodySize: e.encodedBodySize,
    decodedBodySize: e.decodedBodySize,
    duration: Math.round(e.duration),
    protocol: e.nextHopProtocol,
  }));
})()

5. Filter & analyze

  • Apply ${input:filter} regex to name if provided.
  • Group by initiatorType (fetch, xmlhttprequest, script, img, css, font, other).
  • Sort by duration desc → top 10 slowest.
  • Sort by transferSize desc → top 10 heaviest.

6. Screenshot (optional)

Resolve filePath to absolute path, create parent dirs, call mcp__chrome__take_screenshot.

Output Format

🧭 URL

${input:url} — audited at <timestamp>

📊 Summary

  • Total requests: N
  • By initiatorType: fetch: N, script: N, css: N, img: N, font: N, ...
  • Filter applied: ${input:filter} or "none"

🐌 Slowest requests (top 10)

duration (ms)transferSizeinitiatorTypename (truncated)

📦 Heaviest by size (top 10)

transferSizeencodedBodySizeprotocolname (truncated)

🖼 Screenshot

Saved: <absolute path> (if captured)

📋 Raw table (filtered)

Compact table of all matched requests: initiatorType | transferSize | duration | name

Notes

  • transferSize = 0 usually means served from cache (304 or disk cache).
  • HTTP/1.1 connections with many requests → consider HTTP/2 or bundling.
  • Headers and status codes are not available from Performance entries; for those, use targeted fetch() calls via mcp__chrome__evaluate_script.
  • For SPA flows, navigate to the route first, then clear timings before triggering the specific action.

Common Mistakes

  • Forgetting to clear timings first — entries from previous navigations pollute results.
  • Too short a duration — lazy-loaded resources and deferred API calls may not appear.
  • Treating transferSize=0 as missing — it means cached, not absent.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.