Single sheaf modelNo cross-sheaf bridges for this demo yet.
readyCategorical Lens
Read this demo as local sections, restriction maps, gluing, obstructions, and Kan-style structured completion.
Collective intelligence: many agents have only local views, but global structure emerges through compatible local actions.
Objects
sales, inventory, staffing, pricing, customer-experience charts
sales, inventory, staffing, pricing, customer-experience charts
Morphisms
restriction maps over demand-inventory, demand-staffing, pricing-customer, and inventory-customer overlaps
restriction maps over demand-inventory, demand-staffing, pricing-customer, and inventory-customer overlaps
Sections
local operational states and truth values
local operational states and truth values
Gluing
claims or artifacts admitted when local sections agree on overlaps
claims or artifacts admitted when local sections agree on overlaps
Obstructions
qualified, review, or quarantine cells where gluing fails
qualified, review, or quarantine cells where gluing fails
Kan View
manual baseline for extending local store state into global operational guidance
manual baseline for extending local store state into global operational guidance
Prometheus placeholders
TICKET Transformer
TICKET(storefront_static_baseline_run) - Topos Integration using Causal Kan Extension Transformers
Expression: TICKET(storefront_static_baseline_run) -> compose(operational_decision, store_context, inventory_staffing_pricing) -> glue(storefront_admission)
FSQL: FROM prometheus_runs SLICE BY foundry("storefront") TICKET BY target_foundry
Admission policy: quarantine actions when demand-inventory, demand-staffing, pricing-customer, or inventory-customer overlaps fail to glue
Scylla decision console
Storefront foundation model
Store operations are mostly coherent, with pricing decisions needing validation.
5decision regions
4consistency checks
1blocked decisions
What Scylla Can Answer
- Where are demand, inventory, staffing, pricing, and customer experience aligned?
- Which operational decisions can be trusted now?
- Which signals should trigger the next autonomous refresh?
Current Readiness
Needs evidence before action- Update cadence
- hourly operational refresh
- Latest signals
- new sales, staffing, inventory, and customer-experience traces
- Next update
- refresh demand/inventory overlap after the next sales window
- Blocked by
- pricing_customer
Signals Scylla Watches
sales movementinventory pressurestaffing capacitycustomer experience
Integrated Sheaves
Generated sheafSingle-cover Odyssey demo
This demo uses one generated sheaf cover.
Cross-Sheaf Bridges
Recommended Next Actions
- Refresh demand and inventory after the next sales window.
- Hold promotion changes until customer response evidence is stronger.
- Route staffing adjustments through the validated demand/staffing overlap.
Decision Guardrail
Do not trust promotion recommendations until pricing and customer evidence agree.
Evidence Layer
Scylla keeps the user-facing view concise. Prometheus remains available when the internal audit is needed.