Organization, Identity, Access, and Governance — Technical reference

This reference owns the implementation contracts of BND-ORG. Read the package guide for usage, configuration, dependencies and upgrades. Generic execution, isolation and admission remain governed by Platform Documentation.

Data model and interfaces

DataModelPackage: DM-ORG

Owned entities: organization, workspace, business unit, principal, group, role assignment, access review, purpose, acceptance, jurisdiction, retention, legal hold.

InterfacePackage contributions: IF-ORG-ADMIN.

Each business object and source reference remains tenant- and purpose-scoped. The package owns schema evolution and declared migration behavior; the platform enforces schema registration, authorization and versioned installation. Consumers reference this model rather than creating competing canonical copies.

Workflows

WorkflowPackage: WFP-ORG-GOVERNANCE

• Tenant Activation
• Identity Provisioning
• Role Assignment
• Access Review
• Eligibility and Acceptance
• Policy Change
• Retention and Legal Hold
• Exceptional Access Review

These are named definitions within the workflow package, not separate invented workflow IDs. Workflow runs resolve package versions, actor and purpose, authorized inputs, retries, review gates and output schemas through the environment lock. Temporal provides durable execution and Kafka carries declared lifecycle/domain events.

Policies, evaluations and tools

POL-ORG-AUTHORIZATION, POL-PURPOSE-ELIGIBILITY, POL-RETENTION, EVAL-POLICY-CONFORMANCE, identity and directory connectors.

Qualification covers the selected model/runtime profile, languages, access scope, failure handling and the package's domain-specific outcomes. Tool execution uses the platform gateway; package code cannot increase its own permissions or write directly to external systems.

Owned agents

P01 — Context Assembly Agent

Agent ID: effority.context-assembly

Purpose: Builds the minimum permission-scoped evidence and business-context package required for a declared agent purpose.

Triggers: Any authorized agent invocation, context refresh, source correction, or permission change.

Inputs: Policy decision, actor, tenant, purpose, subject references, source versions, time range, language, sensitivity ceiling, and context budget.

Outputs: context-package.v2, selected evidence references, graph slice, taxonomies, explicit exclusions, source authority, and remaining budget.

Tools: evidence.search, transcript.read, document.read, graph.read, finding.read, and policy.explain.

Review and limits: It never interprets business meaning, creates a finding, or reveals restricted-source identity. Authorization occurs before candidate retrieval.

Primary metrics: Retrieval coverage, irrelevant-context rate, ACL violations, context size, latency, and downstream citation use.

Implementation profiles

The package composes the data, workflows, policies and interfaces above using the shared runtime. The owned-agent contracts specify inputs, outputs, tool boundaries, review requirements and measurable outcomes. Deterministic platform operations are not represented as additional agents.

Integration boundaries

Identity enforcement stays in the platform. A business role never automatically grants access to private coaching or restricted ethics cases.

Owned screen contracts: SCR-ADM-01, SCR-ADM-02, SCR-ADM-04.

See solution workflows for cross-package behavior. Screens, agents and workflows keep one canonical owner even when other bundles invoke or enrich them.