pack
Site Analytics Engine
A weekly growth report from GA4, GSC and Stripe — ranked fixes, not dashboards.
✓ = we've run it on that platform · others load the same SKILL.md format but aren't run-tested yet
The engine behind AgentKit Works' own Monday board reports. Three fully implemented collectors — GA4 traffic, Search Console queries, Stripe revenue — feed one analysis pass that returns 3–5 ranked recommendations with the data behind each, then drafts the safe fixes for your approval. Adapter patterns for four more sources (email, distribution, AI citations, catalog health) show your agent how to wire in whatever your stack has. Built to run as a Monday-morning cron.
Why it's different
- ▸ Analysis engine turns raw data into 3–5 ranked recommendations, each with the number behind it
- ▸ Drafts safe improvements for your approval — stats blocks, changelog entries, copy fixes
- ▸ Any unconfigured source is skipped and named in the report — never faked, never a crash
- ▸ Production-tested: writes AgentKit Works' own weekly board reports
Questions, answered
How long to set up?+
Stripe is immediate (one restricted key). GA4 + GSC take the guide's 10-minute path — one Google Cloud service account covers both. Optional report history needs a free Supabase project (two env vars); without it, reports just land in git.
Does this work without GA4?+
Yes. Every source degrades gracefully and the report names what it skipped. Start with Stripe for revenue, add GA4 and GSC when you want traffic and search data.
How is this different from Distribution Autopilot's weekly-analytics play?+
That play is one of nine inside a publishing system and reads the snapshots its own collectors write. This kit is the standalone, deeper version: its own collectors, a ranked-recommendation engine, and drafted fixes — no publishing machinery required. If you run both, this kit's report is the Monday read and Distribution Autopilot acts on it.
Kits that pair with this one
Each workflow kit teaches a different job. They share the same skill standard, so they stack.
Prompt Ops: 150 Production Prompts
150 production prompts, annotated — the failure modes documented for every one.
$29$19
VIEW
Skill Forge
43 production agent skills — plus the framework and test harness to write your own.
$89$59
VIEW
Outreach Machine
Cold outreach your agent researches and drafts — enrichment → sequence → reply triage.
$69$49
VIEW