SHIPPED-0.1 / SHIPPING

The Kickstarter-to-Shipped Benchmark

How much missing fulfillment judgment can each workflow supply from one request?

RELEASE SHIPPED-0.121 JUL 2026MODELED

STATUS: Modeled · no controlled runs
INPUT: One prompt
PRIMARY: Operational judgment
OUTCOME PACK: Kickstarter Fulfillment ↗

01 / ONE PROMPT

Use the request a person would actually make.

USER INTENT
“Get this funded Kickstarter shipped.”

Get this funded Kickstarter shipped./goal Get this funded Kickstarter shipped.$possible Get this funded Kickstarter shipped.

02 / STARTING POINT

Each workflow starts with the same subject.

SAME STARTING POINT

A funded campaign and its reward obligations. No production, order, logistics, exception, or backer operation exists yet.

03 / OPERATIONAL JUDGMENT

What did each workflow know to add?

Scores credit visible evidence only. They do not measure artifact quality.

Operational judgment supplied by Direct, /goal, and $possible
Missing expertise	Direct	/goal	$possible
Infers omitted operational scope	PARTIALShipping plan only`shipping-plan.md`	PARTIALProduction, orders, exceptions`production-plan.md · orders.csv · exceptions.md`	EXPLICITObligations through shipment evidence`fulfillment/campaign-obligations.md · fulfillment/production/plan.md · fulfillment/quality/acceptance.md · fulfillment/inventory/state.csv · fulfillment/logistics/plan.md · fulfillment/milestones/shipment-ledger.csv`
Separates evidence from assumptions	NONENo visible artifact	NONENo visible artifact	PARTIALObligations and shipment evidence; no assumptions register`fulfillment/campaign-obligations.md · fulfillment/milestones/shipment-ledger.csv`
Surfaces risk and safeguards	NONENo visible artifact	PARTIALException document`exceptions.md`	EXPLICITQuality, privacy, exception controls`fulfillment/quality/acceptance.md · fulfillment/privacy/boundary.md · fulfillment/exceptions/queue.csv`
Defines acceptance or decision gates	NONENo visible artifact	NONENo visible artifact	PARTIALQuality acceptance, no visible action gate`fulfillment/quality/acceptance.md`
Creates durable operating state	NONENo visible artifact	PARTIALOrder ledger only`orders.csv`	EXPLICITOrders, inventory, exceptions, control, shipments`fulfillment/orders/ledger.csv · fulfillment/inventory/state.csv · fulfillment/exceptions/queue.csv · fulfillment/control/state.json · fulfillment/milestones/shipment-ledger.csv`
Preserves review and completion evidence	NONENo visible artifact	PARTIALGeneric result record`RESULT.md`	EXPLICITFirst-cycle + outcome receipts`fulfillment/receipts/first-cycle.md · fulfillment/outcome-receipt.md`

04 / ARTIFACTS AND TIME

Artifacts and time support the comparison.

0918 artifacts

Direct2
AGENT RUNTIME 14 minHUMAN COORDINATION 1 min
Produces a high-level shipping plan.
/goal7
AGENT RUNTIME 47 minHUMAN COORDINATION 2 min
Adds milestones, ownership, and a completion record.
$possible18
AGENT RUNTIME 5 hr 20 minHUMAN COORDINATION 3 min
Compiles the production-to-95%-shipped operation.

Direct2 ARTIFACTS

shipping-plan.md
backer-update.md

/goal7 ARTIFACTS

GOAL.md
production-plan.md
orders.csv
shipping-plan.md
backer-updates/
exceptions.md
RESULT.md

$possible18 ARTIFACTS

fulfillment/campaign-obligations.md
fulfillment/production/plan.md
fulfillment/quality/acceptance.md
fulfillment/suppliers/dependencies.md
fulfillment/backers/schema.md
fulfillment/orders/ledger.csv
fulfillment/privacy/boundary.md
fulfillment/inventory/state.csv
fulfillment/logistics/plan.md
fulfillment/exceptions/queue.csv
fulfillment/communications/cadence.md
fulfillment/updates/drafts.md
fulfillment/support/playbook.md
fulfillment/control/state.json
fulfillment/receipts/first-cycle.md
fulfillment/schedule/task.md
fulfillment/milestones/shipment-ledger.csv
fulfillment/outcome-receipt.md

05 / WHY THE GAP EXISTS

Possible starts with a reviewed outcome map.

Direct and /goal infer the missing work. $possible compiles it from the matching pack.

Inspect the Kickstarter Fulfillment outcome pack →

← All benchmarks02 / 02