agentleFS
itsryanarmaniitsryanarmani/skills/ad-landing-tracking/SKILL.md

ad-landing-tracking

Skill1 starsChanged 9 days agoMITView on GitHub

Track a paid ad click from landing page to payment: click-id and UTM allowlist, first-touch and latest-click cookies, ids carried through checkout into payment metadata, one shared event id for browser and server dedup, variant pages keyed by a fixed slug, UTM naming, safe reports, dry-run spend scripts, and a paused-first launch. Use when you build or audit ad landing pages, attribution capture, or a small paid test, and conversions do not match clicks.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. About 0 are common, 14 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈14

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not use it to pick a platform or write ad copy.

    line 18
  2. Distinctive:

    Write cookies on the first touch and on new or changed params only, after consent where the law needs it.

    line 23
  3. Distinctive:

    Read ids at the purchase click.

    line 24
  4. Distinctive:

    Use one event id in the browser and on the server.

    line 25
  5. Distinctive:

    Run every spend or upload script as a dry run first.

    line 27
  6. Distinctive:

    Carry: Put the allowlisted params on the CTA link, in storage, and into checkout metadata.

    line 33
  7. Distinctive:

    Read: Use one metric window for all channels.

    line 36
  8. Distinctive:

    Launch: Follow the checklist below.

    line 37
  9. Distinctive:

    Prove: Check pixel and server events on the live page with a clean browser.

    line 38
  10. Distinctive:

    Stop at the first item that fails.

    line 44
  11. Distinctive:

    Check the spend cap. Use a lifetime or total cap where the platform has one.

    line 49
  12. Distinctive:

    Create the ad paused. Read back targeting, link and budget.

    line 50
  13. Distinctive:

    Wait for the stop rule's minimum clicks.

    line 51
  14. Distinctive:

    Read the first result on one metric window.

    line 52

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

ad-landing-tracking

Capture every click id once, keep it until the sale, and send it back to the ad platform with one shared event id.

This skill covers the page and the pipe. For the event taxonomy and platform adapters, use ad-conversion-hub. For hypothesis design and budgets, use ad-experiments.

Use it when

  • You build a landing page for paid traffic.
  • Platform conversions are lower than your own counts.
  • You add a variant page or a new platform to a working setup.

Do not use it to pick a platform or write ad copy.

Rules that never bend

  • Allowlist the params you keep. Drop the rest.
  • Write cookies on the first touch and on new or changed params only, after consent where the law needs it.
  • Read ids at the purchase click. Do not stamp them at signup.
  • Use one event id in the browser and on the server. Make it once.
  • Show unknown as unknown. Never show it as zero.
  • Run every spend or upload script as a dry run first.
  • Launch every ad paused. A person activates it.

Workflow

  1. Capture. Keep click ids and UTM fields in two lists. Build one allowlist from both. Keep first touch for source reports and the latest Meta and Reddit click id for sending. Write cookies only after consent where the law needs it. Never store the raw query string. Details: identifiers-and-cookies.md.
  2. Carry. Put the allowlisted params on the CTA link, in storage, and into checkout metadata. Read them back in the payment webhook. Details: same file.
  3. Send. Make one event id once: derive it from a stable key, or create it in the browser and send it to the server. Use each platform's exact field name. Details: events-and-dedup.md.
  4. Vary. Serve each variant at a fixed slug. Put the arm in the UTM fields and a variant param. Details: variants-and-naming.md.
  5. Read. Use one metric window for all channels. Keep the last good report. Details: reporting-and-launch.md.
  6. Launch. Follow the checklist below.
  7. Prove. Check pixel and server events on the live page with a clean browser. An automated browser sends nothing by default. Details: verify-events-live.md.

For page rules (canonical, no header or footer, catalog-built copy, no paid content), see ad-landing-pages.md.

Launch checklist for a small test

Stop at the first item that fails.

  1. Fire a test event from the browser and one from the server. Use the platform's test id. Confirm both appear and merge into one event.
  2. Click a live ad URL. Confirm the first-touch cookie holds the UTM fields and the latest-click cookie holds the click id.
  3. Buy through checkout with a test card. Confirm the webhook reads the click id from the payment metadata.
  4. Check the spend cap. Use a lifetime or total cap where the platform has one. Otherwise set a daily budget times the days, with an end date. Also set a hard account spend limit. A daily budget alone is not a cap.
  5. Create the ad paused. Read back targeting, link and budget. Compare with your plan. Activate only if all match.
  6. Wait for the stop rule's minimum clicks. Do not read results before that.
  7. Read the first result on one metric window. Check unknown cells before you read the zeros.

Failure notes

SymptomLikely causeFix
Click id missing after a redirectThe redirect response did not set the cookieSet the cookie on every response path
Events count twice in the platformEvent ids differ, or a field name has the wrong caseUse the shared id and the exact name per side
Fast bounces never appear in analyticsFirst page view fires after hydrationSend it from the inline init script
Page shows {{token}}A token had no valueReplace it with a default or empty text
Leads vanish with no errorLogin check covers the lead APIExempt the public lead API and test as anonymous
Report went blankEvery row failed to measureKeep the last good report
Zero pixel events in an automated checknavigator.webdriver is true or a blocker runsUse the clean check in verify-events-live.md
Lead counts twiceEvent names or ids differ, or the server sends Lead on repeat sign-upsSame name and id on both sides; send Lead on first sign-up only

What we could not find

The source review found no sample-size rule, no power analysis, no significance test and no bandit. The only stop rule is a floor on spend, conversions and return. It is not a test of significance.

For measurement design, see the measurement guide when it lands.

Unproven here: page speed budgets, cache lifetimes for ad pages, and any ranking of page structures.

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.