AdMapix Raw Data Developer Helper

SKILL.md

---
name: software-data-admapix-raw-developer-helper
description: >-
  Design and harden AdMapix-style raw data workflows for creative, app, ranking, revenue, and analytics datasets before they become dashboards or agent tools. Use when a user asks for AdMapix, raw data, app analytics, creative data, rankings, or needs practical workflow, code, checklist, documentation, or review support for this job.
---

# AdMapix Raw Data Developer Helper

## Purpose

Use this skill when a user needs to ingest, normalize, inspect, or expose raw advertising/app-market data while keeping provenance, schema drift, deduplication, and downstream safety under control.

Audience: data engineers, growth analysts, skill authors, and agent builders adapting popular AdMapix-style raw data workflows.

Read `references/requirement-plan.md` when demand evidence, source links, scoring rationale, or review criteria are needed.

## Workflow

1. Identify raw sources, refresh cadence, entity keys, metrics, dimensions, privacy constraints, and target consumers.
2. Create a source-to-table map that separates immutable raw captures from cleaned views, feature extracts, and presentation layers.
3. Specify validation rules for required columns, date ranges, currency, locale, creative IDs, app IDs, duplicate rows, and missing metrics.
4. Plan failure handling for API quota, partial exports, schema drift, backfill gaps, and vendor naming changes.
5. Recommend lightweight local implementation steps using CSV, JSON, SQLite, or DuckDB before heavier warehouse work.
6. Return an audit trail with source, transform version, row counts, anomalies, and consumer-facing caveats.

## Expected Outputs

- A raw-to-clean data model for AdMapix-style inputs.
- Validation checks for schema drift, duplicates, missing values, and metric consistency.
- A local script or pseudocode plan for ingestion and normalization.
- A reliability checklist for publishing the workflow as an agent skill.

## Validation

- Raw data remains reproducible and separate from cleaned outputs.
- Every transform records source, timestamp, row count, and known anomalies.
- The workflow handles partial data, schema drift, and quota failures explicitly.
- The recommended implementation can run on ordinary CPU hardware.

## Triggers

Keywords: `AdMapix`, `raw data`, `app analytics`, `creative data`, `rankings`, `revenue`, `ETL`, `data quality`

Example trigger sentences:

- `Use $software-data-admapix-raw-developer-helper to design a raw data layer for ad creative exports.`
- `Clean this app-ranking CSV without losing the original source data.`
- `Create validation checks for an AdMapix-style analytics skill.`