agentleFS
sergeyizmailovsergeyizmailov/knowledge-delta-skills/skills/ads-core/tracker-ops/SKILL.md

tracker-ops

Skill3 starsChanged 49 days agoMITView on GitHub

Tracker ops for affiliate/media buying: Keitaro + Binom APIs (keys, reports, cost push), postback/S2S, metric discipline (payout event vs all-conversions, timezones, CPL math), daily spend-sync, gclid/OCI chain for Google. Principles port to other trackers.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    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:

    No cost push = no CPL/ROI (report cost is 0 — or fake: CPA/CPS cost model with auto ON adds fixed cost per conversion, 01).

    line 55

≈ “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

Tracker Ops

Count money correctly and automate tracker work. Keitaro + Binom specifics below; the metric rules and daily sync port to any tracker. Baseline 2026-09-03 (Sonnet 5 blind, CPL cohort task): [R] — click-date cohort and pinned attribution window already known; do not expand that section.

Metric rule (NEVER violate)

  1. "conversions" (Keitaro) can exceed your payout-lead count, but that is INTEGRATION-SPECIFIC, not an inherent 3.5-4x: the same subid+status OVERWRITES; extra records come only from distinct tids or a genuine multi-step funnel. Never do CPL math on "conversions" — the payout metric is ONE event/status; get it from the TL in writing. (Any "3.5-4x" is one setup's ratio, not a law.) Exception: statuses flip in place, so a REG payout counts leads + sales (rule 2).
  2. Metric names aren't portable: Keitaro leads = conversions STILL in status lead (lead→sale flips in place): regs = leads + sales, deps = sales — leads as reg count overstates reg→dep (FIELD 2026-09-27: 11 vs 17). Binom "leads" ≈ all conversions. Verify semantics per tracker per setup.
  3. Cross-check: platform-reported lead count ≈ tracker lead count. A big divergence = wrong metric or broken tracking; stop and reconcile before reporting. The "±20%" tolerance is an account-specific baseline you set from a reconciled day, not a universal constant. Meta counts view-through the tracker never sees — compare Meta click-only (03).
  4. Compute daily CPL in the AD ACCOUNT timezone for both spend and leads. Google's account timezone is permanent (senior-buyer-ops contract #5) — reconcile in the tracker, never by eye.
  5. Rows with unsubstituted {{...}} macros: exclude from performance analysis, then classify the cause — bots/crawlers, yes, but also a broken URL template, an unsupported/misspelled macro, or manual/direct traffic. Fix the template if it's yours before blaming traffic.
  6. Money = sale_revenue, never revenue (includes lead_revenue; payout on lead postbacks fakes ROI — FIELD 2026-09-27: $9,820 lead_revenue vs $1,650 real sale_revenue). Check lead_revenue = 0.

Route references

NeedReference
Keitaro: key, report/build, measures, update_costs, postback/S2S, postback drill, gotchasreferences/01-keitaro.md
Binom: legacy URL-param + v2 REST, reports, cost, postback/S2Sreferences/02-binom.md
CPL math, funnel + anti-fraud metrics, cohort nowcasting, backend optimization contract (status→CAPI event), TMA/bot CAPI without pixel, multi-tracker sync + conversion ledger, ATT asymmetry, mapping, daily routinereferences/03-metrics-and-math.md
Google lane: gclid/gbraid/wbraid, ValueTrack chains, Keitaro/RedTrack Google config, OCI windows + dedup, the 2026-06-15 cutoff, Data Manager events:ingest for new tokensreferences/04-google-lane.md

Platform lanes are not interchangeable — Meta counts through CAPI, Google through gclid → OCI, with different windows, a different dedup key, and a hard onboarding cutoff. Read 04 before wiring anything on Google traffic.

Non-negotiables

  • Work only on YOUR campaign IDs; every call carries a campaign filter. Other buyers' campaigns are read-never-touch.
  • No cost push = no CPL/ROI (report cost is 0 — or fake: CPA/CPS cost model with auto ON adds fixed cost per conversion, 01). Push daily, keyed on ids (ad_campaign_id / sub_id_6), in the ad account's tz and currency: metaops keitaro push (dry run first, then --confirm PUSH; 01). Overwriting a period is idempotent, but cost lands only on clicks matching range+filters — reconcile source spend vs tracker cost after each push.
  • API keys are write-once; store verbatim in gitignored notes.

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.