A decision-to-outcome control plane: turn an uncertain signal into an evidence-backed commitment, govern its execution, measure the result, and retain the record.
docs/user-guide/.Up to seven destinations across the top — each an honest view over the same records. Portfolios, Records and Admin are hidden in a personal workspace (decluttering, not a permission).
Home
Your items, their real lifecycle stage, and a count of what’s waiting on you. No invented “velocity” charts.
For You
The universal inbox — a live query of everything that needs you. No Dismiss: items leave when you act on their source.
Portfolios
TeamPrograms, policies, investments and initiatives, grouped by type.
Work
Every strategic item in the workspace — the searchable list, and the front door to each item workspace.
Intelligence
Cross-item signals computed from rows that exist — evidence health, gate pressure. Never a model’s opinion.
Records
TeamOrg-wide decisions, approvals, artifacts and audit. Append-only — nothing here can be edited or removed.
Admin
TeamA directory of admin surfaces, each enforcing its own access. Reaching it doesn’t make you an admin.
One strategic item, six destinations, one lifecycle stage. Open a stage to see the destinations it covers and the exact question each answers.
DeciSense refuses to show precision it doesn’t have. Here’s what is admin-only, preview, or not-yet-wired — with the tracking issue.
Evidence links
Citations show as “unverifiable” — the per-reader permission evaluator isn’t built yet (ISS-020). The evidence is shown; the link is withheld.
Outcomes & objectives
Schema and domain logic shipped; the create/write path through the UI is a tracked follow-up (ISS-024). Read and structure today, author soon.
Connectors & webhooks
You trigger a sync and register a webhook. Continuous background sync and live delivery are built but not drained by a running worker yet (ISS-032).
Tenant isolation
Enforced by the application today — every query scopes to your org and workspaces. Database row-level security is proven but not yet live (ISS-027).
SSO / SCIM
Configurable now; the IdP handshake is mock-wired pending your credentials, and SAML awaits a licence-approved library (ISS-031).
AI usage policy
Authored and decidable; enforcing it at the live LLM call site is a follow-up (ISS-029). On air-gap the egress guard already guarantees no inference leaves the boundary.