// guides + reports

Long-form playbooks for autonomous commerce.

Guides, reports, and playbooks written by the team building Magistry and the operators running it. Each one is the reference doc we wished existed when we hit the same wall.

Read the 2026 report

Browse all guides

9 long-form pieces 5 categories No email gate on guides

// filter

Browse by category.

// featured report

The 2026 Autonomous Commerce Report.

Released this month. 48 pages. Free, no email gate — read the whole thing in the browser.

REPORT

The 2026 Autonomous Commerce Report.

A practical field guide to shipping autonomous agents on Shopify — the patterns and metrics that matter.

// pages

Field guide

// contributors

Magistry

// released

2026

Read the report Browse all guides

// table of contents

Our perspective on running autonomous agents against a live store: the dry-run timeline that works, the kill-switch drills worth doing, self-calibrating thresholds versus rules of thumb, and the decision_log queries a CFO will ask for. Analysis and worked examples — not a customer survey.

›01 · The state of autonomous commerce in 2026
›02 · Dry-run to Phase 2 — the 14-day pattern
›03 · Reading the decision_log: queries that matter
›04 · Kill switch drills: monthly cadence, quarterly audit
›05 · Self-calibrating thresholds vs. industry rules of thumb
›06 · The CS reply judge — where to set the floor
›07 · Cost-confidence tiers and what they protect
›08 · 2027 outlook: what's next after autonomy

// all guides

Eight more playbooks from the bench.

Guides explain. Playbooks prescribe. Reports survey. Every piece below is read-time-honest — if it says 30 minutes, it's 30 minutes of real reading.

GUIDECatalog

Catalog audit in 10 days.

A day-by-day plan for taking a 1,000+ SKU store from 'unsorted Shopify export' to a tier-stamped, lifecycle-classified catalog ready for the agents.

22 min readRead

PLAYBOOKCampaign

The Customer Match audience playbook.

How to seed Google Customer Match from your decision_log — purchaser cohorts, refund-excluded segments, and the LTV bands that scale.

18 min readRead

GUIDECS

Reply Judge — what to gate, what to ship.

How the multi-dimension gates work — policy grounding, factual grounding, brand voice — what to always route to a human, and how to drift-detect monthly.

14 min readRead

GUIDEAudit

decision_log — a 30-minute primer.

The schema, the bitemporal versioning, the five queries every operator runs in the first week, and the rollback patterns that build trust with finance.

30 min readRead

PLAYBOOKOperations

Setting up your first kill switch.

Who holds it, what triggers it, what it actually stops, and the monthly drill that keeps the muscle memory alive.

12 min readRead

PLAYBOOKOperations

From dry-run to live in 14 days.

The exact ramp: day 1 connect, day 2 first dry cycle, day 7 review, day 14 Phase 2 flip — with the per-action rate-limit table we recommend out of the box.

20 min readRead

GUIDECatalog

Cost confidence tiers explained.

Tier A, B, and C — what they mean, what each one unlocks, and how the margin policy stays enforced in the type system rather than a Notion doc.

11 min readRead

PLAYBOOKOperations

Magistry for multi-store agencies.

Workspace structure, per-client kill switches, operator role mapping, and the audit pack template that wins the renewal conversation.

26 min readRead

// the bar

Three editorial rules we won't break.

No SEO bait.

If the post wouldn't earn a Sunday morning of an operator's time, we don't ship it. Topic clusters are an output, not a strategy.

Real read times.

Measured at 230 words per minute. If a guide says 22 minutes, it's because we timed it ourselves end-to-end.

No email gate.

Guides and reports are free to read. The dispatch is opt-in. We'd rather lose the lead than rent it.

// from theory to write

Want a guide written about your store?

We'll ghost-write the 'how we shipped Phase 2 in 14 days' case for any operator who flips Phase 2 inside their first month. Your data, your prose, our editorial bar.

Book a 20-min demo

See past cases

Dry-run by default · Append-only logs · One-click rollback