ga4-pack
Investigate GA4 report discrepancies, realtime event observations and BigQuery export gaps using evidence receipts, executable analyzers and bounded optional collection.
Installation
Open Claude Code and run this command:
/plugin install ga4-pack@claude-code-plugins-plus
Use --global to install for all projects, or --project for current project only.
What It Does
Investigate conflicting reports, missing measurement signals and BigQuery export gaps with executable evidence checks. The v2 rebuild retains the ga4-pack install identity and consolidates five tutorials into three operator workflows.
Install: /plugin install ga4-pack@claude-code-plugins-plus
Freshie run 139 grades the three workflows A/92, A/95 and A/97. See the value decision and validation evidence for the immutable source, public inventory receipt and verification limits.
| Workflow | Concrete output |
|---|---|
| Report Forensics | A comparison ledger separating scope, identity, quality and row-coverage differences from hypotheses. |
| Measurement Verifier | Expected-event observations with count, truncation and deployment-window limits. |
| Export Guardian | Daily/intraday evidence and gaps, plus an optional SQL-bound dry-run risk assessment. |
Each workflow has a stdlib analyzer, a bounded optional SDK collector, a receipt contract, primary references and behavioral evaluation cases. Supplied-data mode works without credentials. Live collection requires an explicit live invocation, an approved target and existing credentials. Development checks use synthetic transports; they do not establish live property access or model-evaluation results.
Python 3.10+ is required; historical/export analysis also needs IANA timezone data. Optional collectors use google-analytics-data or google-cloud-bigquery from an approved environment. No key creation, IAM grants, tracking changes, export setup or paid query execution is performed. A successful dry run does not approve spend. Keep analytics receipts and SQL in approved local storage.
Skills (3) plugin-local skills
Investigate GA4 BigQuery export gaps, daily/intraday overlap and query scan risk from table-inventory and dry-run receipts.
Check expected GA4 event names and counts against a realtime report receipt, preserving row coverage, minute-window and deployment-timing limits.
Investigate conflicting GA4 report numbers using request and response receipts.