The Reporting Engine
Skills · Subagent · Command · PDF
.claude/the-reporting-engine/zip
skills/06
├ anomaly-triage
├ kpi-dictionary
├ pacing-monitor
├ report-narrative
├ report-qa
├ weekly-digest
agents/01
├ data-auditor
commands/01
├ /weekly-digest
FIELD-GUIDE.pdf8pp
English
One payment · No account · No subscription
- ✓ Download link in your inbox in seconds
- ✓ Secure card checkout · VAT handled
- ✓ 14-day money back, no questions
- ✓ Every future revision included
- ✓ Runs in Claude Code and claude.ai
Built for Claude first, our recommended agent. Skills follow the open Agent Skills standard (SKILL.md), so they also run in any agent that reads it; subagents and slash commands are Claude Code formats.
The Reporting Engine
Raw exports in, defensible numbers out
six skills·one subagent·one command·8-page guide
Raw platform exports in, a client-ready week out: metrics normalized before they are summed, drops triaged against the flight plan, every number checked before it is sent.
Executive summary
- The problem
- Meta, Google Ads and GA4 each define a conversion differently, so the total in a weekly report is often a sum of things that cannot be added, and a dip that was only a flight ending reaches the client as a crisis. It looks right, so nobody checks until the client does.
- What it does
- Six skills and a read-only auditor normalize the exports before anything is summed, triage every drop against the flight plan, and re-derive each figure from the source files before the report is sent.
- Why take it
- Your name is on the report. This makes every number in it one you can defend, every week, whoever on the team writes it.
For analytics leads and the marketer who owns the Monday report.
In the first week you will
- 01Run /weekly-digest on Monday's exports and send the client summary in under 250 words and one table
- 02Normalize Meta, Google Ads and GA4 columns with kpi-dictionary before anything is added together
- 03Send the draft and its source files to the data-auditor subagent and get the mismatches before the report leaves
Contents of the Kit
- 01Six installable agent skills: kpi-dictionary, anomaly-triage, pacing-monitor, weekly-digest, report-narrative, report-qa
- 02The data-auditor subagent: re-derives every number in a draft report from the source files and lists the mismatches
- 03A /weekly-digest slash command for the Monday summary
- 04The cross-platform metric dictionary: how Meta, Google, DV360, TikTok, Pinterest, LinkedIn and GA4 define spend, reach, views and conversions, and which of them may be added together
- 05An 8-page Field Guide on the reporting chain and the checks a human still does
- 06Plugin manifest included: install as a Claude Code plugin or copy the folders
One skill, as it ships
SKILL.md
kpi-dictionary
- Use when
- Use when consolidating exports from two or more platforms, when a column name is ambiguous (Results, Revenue, Views, Clicks), before building any cross-platform table, or when a client asks why two platforms disagree.
- Requires
- Exports
- Will not
- No column is mapped by name similarity alone · Do not invent a definition for a platform outside the table · Reach and platform-attributed conversions are never summed across platforms
- Returns
- KPI Dictionary: {client or account set} / {period}
One of six skills in the kit · All of them install the same way
The contents, in full · six skills, one subagent, one command
- 01kpi-dictionaryThe metric normalization step. Before anything is summed, each metric is classified as comparable, comparable with a caveat, or not comparable across the platforms in the export, and the skill carries the reference table so the classification does not depend on who is on shift.
- 02anomaly-triageSeparates real regressions from noise: paused-by-design campaigns, flight ends, budget caps, delayed attribution, reporting lag, tracking breaks. Produces a triage log with severity, likely cause, evidence and the next check, and a rule for when a client hears about it.
- 03pacing-monitorBudget against spend to date, projected end-of-flight spend, days ahead or behind, recommended daily adjustments. Every projection is labeled linear or seasonality-adjusted so nobody mistakes an assumption for a forecast.
- 04weekly-digestThe client-ready Monday summary: what moved, why, what happens next week, under 250 words and one table.
- 05report-narrativeThe words on top of a monthly, quarterly or campaign report. Observation, explanation and decision stay in separate registers, a cause is only written when it is in a log with a date and a stated mechanism, and every decision carries an owner, a review date and the figure that would reverse it.
- 06report-qaThe pre-send checklist: totals reconcile with platform totals, date ranges match, missing days are found, currencies are consistent, month-on-month math holds, no non-comparable metric is compared, and every claim in the narrative traces to a number.
- 07data-auditor (subagent)A read-only second pair of eyes. It re-derives every figure in a draft report from the source files and reports what does not match.
- 08/weekly-digestThe slash command for the Monday summary, run on this week's normalized numbers, pacing table and triage log.
Most reporting time is not spent on insight. It is spent on the steps before it: mapping columns, reconciling totals, and working out whether a drop is a regression or a flight that ended on schedule. The kit follows one rule from the manuals: nothing writes a sentence about a number until the number has been normalized and checked against the source.
Pairs with
Think Like a CMO explains what a marketing leader is accountable for; this kit builds the reporting that proves it. The Augmented Practitioner explains the method; the Reporting Engine runs it.
Before you buy
01Do I need Claude Code?
No. Every skill uploads to claude.ai as a zip. Subagents and slash commands are the Claude Code extras.
02I am not a programmer.
Install is copying folders into a .claude directory. The Field Guide in each kit assumes nothing beyond that.
03Will it work in my agent?
Skills follow the open Agent Skills standard (SKILL.md), so any agent that reads it can run them. Built for Claude first, our recommended agent.
04Do the tools invent numbers?
Never. Every figure is supplied by you, confirmed by you, or labeled as an assumption in the output.
05How does it arrive?
Instantly. Pay by card and the download link is in your inbox within seconds, and on the confirmation screen.
06Is this a subscription?
No. One payment, lifetime access to your download, every future revision of what you bought included.
07What if it is not useful to me?
Write to info@conclusionmp.com within 14 days of purchase and you get a full refund. No questions, no forms.
Ready when you are
The Reporting Engine, $12.
Instant download, one payment, every future revision included. Or take The Agency Desk, four items for $35, and this kit works out at $8.75.