Business capabilities

Capabilities describe outcomes the organisation can reliably produce. They are different from teams, roles, activities, or tools: Resolve customer incidents is a capability; a support team, ticket-triage activity, or AI agent may help deliver it.

Plan availability: Capabilities are available on the Enterprise and Agency plans. The Work catalog is available on every plan.

Use the Capabilities workspace to make a transformation proposal explainable:

• define the business outcome once
• assess its current state in Live
• model a different state inside a scenario
• trace the Work and candidate AI solutions that support it
• record outcome measures, evidence, ownership, limitations, and confidence

Availability depends on your organisation's plan, permissions, and current rollout access.

Understand the Live and scenario context

A capability has a stable definition and a contextual assessment:

This means you define Forecast service demand once for the client, assess it in Live, and then change its target maturity or AI readiness in a scenario. Editing the shared name, outcome, hierarchy, or measure definition changes it everywhere in that tenant. Editing the assessment changes only the context shown in the header.

What the capability list shows

The list shows the capabilities assessed in the context selected in the top tabs. In a scenario it also keeps any capability the scenario has retired, so you can see what the target design removes. The first summary card shows how many capabilities are assessed here out of the full shared catalogue.

Capabilities that exist in the shared catalogue but have no assessment in this context are hidden by default. A note above the list says how many there are. Select Show them or tick Show full catalogue to include them; they appear as Not assessed. Open one and use Create Assessment to assess it in the current context. Choosing Not assessed in the Status filter also shows them.

A missing assessment is different from an Absent status. Not assessed means nobody has assessed the capability in this context yet. Absent means it has been assessed and the organisation cannot provide it here.

Snapshots preserve contextual capability records as read-only. Promoting a scenario carries its assessment, mappings, values, and eligible evidence links into the new Live baseline.

Create a capability

1. Open Capabilities from the main navigation.
2. Select New Capability.
3. Give it a short outcome-oriented name.
4. Write the outcome statement in business terms, without naming the team or tool that performs it.
5. Set the first status, maturity, AI readiness, strategic importance, and evidence confidence.
6. Open the new capability to complete its assessment.

The modal separates the two kinds of information. Shared definition (name, outcome, parent, and notes) is added to the catalogue for Live and every scenario. Assessment in the current context applies only to that context; other contexts show the new capability as not assessed until you assess it there.

Capabilities may be grouped under a parent. Keep the hierarchy shallow so it remains useful in client conversations rather than becoming a process taxonomy.

Every field label has an information icon with a short explanation. Maturity, performance, resilience, and AI readiness use a six-point selector from 0 to 5; strategic importance uses 1 to 5. The labels beneath each scale describe its endpoints so scores can be applied consistently.

On wider screens, the creation modal keeps status and evidence confidence side by side, with the three scoring scales on one row. Smaller screens stack the same fields without changing their meaning or requirements.

Required, optional, and Advanced fields

Every capability form labels fields as Required, Optional, or conditionally required. Required fields also have a red asterisk. Advanced keeps useful supporting detail available without making the main task feel longer.

• Capability definition: name and outcome statement are required. Parent capability and description are optional and appear under Advanced.
• Assessment in the current context: status and evidence confidence are required. All five scores, owning team or org unit, accountable position, rationale, and measurement limitation are optional. Ownership and accountability remain visible in the main assessment; rationale and measurement limitation sit under Advanced.
• Supporting Work: activity and contribution type are required for each row. Contribution weight and rationale are optional under Advanced.
• Outcome measure: name, unit type, unit code, desired direction, and aggregation method are required. Description is optional under Advanced.
• Measure value: measure, value type, as-of date, confidence, and source are required. Supply either one point value or both range bounds. Range fields and the optional note are under Advanced.
• Evidence reference: label, evidence type, origin, purpose, permitted audience, retention policy, as-of date, confidence, and approved summary are required. Source system, source owner, and source reference are optional under Advanced.

An Advanced section opens automatically when it contains a validation error, so an invalid field is never hidden from you.

Complete the assessment

The Assessment tab is headed with the active context, for example Assessment in Live, and changes only that context. The Definition tab is marked Live and all scenarios and warns that saving updates the capability everywhere. Archiving a capability asks for confirmation because it removes the capability from the active catalogue in every context. A capability can therefore be operational in Live and scaled in a proposed scenario without duplicating its definition.

Record:

• status: absent, emerging, operational, scaled, or retired
• maturity, performance, resilience, and AI readiness on the 0–5 rubric
• strategic importance on the 1–5 rubric
• owning org unit and accountable position
• evidence confidence and a concise rationale
• a measurement limitation when no defensible outcome value is available

A missing assessment means not assessed. It does not mean the capability is absent.

Organisational ownership and human accountability are deliberately separate. Choose the team, department, or other org unit responsible for sustaining the capability, then optionally choose the human position answerable for its outcome. Occupied positions show the current person's name after the position title; vacant positions remain selectable so accountability survives personnel changes. If the selectors are empty, use the links beneath them to add a team, department, or position in the active Live or scenario context first.

Connect Work and AI solutions

In Work, map activities as core, supporting, or control work. The optional percentage is descriptive contribution weight, not staff allocation or a value formula.

The AI solutions panel shows candidate deployments whose published claims match the categories of mapped Work. This is a discovery aid, not proof that the solution delivers the capability. Use outcome measures and evidence for that claim.

Add outcome measures

Measures stay stable across Live and scenarios; values are contextual. Define the unit and whether higher, lower, or a target range is better, then record dated values as:

• Observed for measured historical or current performance
• Modeled for a scenario estimate
• Target for the intended result

For target-range measures, enter both bounds and leave the point value empty. For other values, enter a point and leave the range fields empty.

Attach evidence safely

Evidence records contain an approved summary, source reference, owner, date, audience, retention policy, and confidence. Orgonaut does not copy the underlying file or raw confidential content into the evidence record.

Use Client audience only when the material may be shown under the engagement. Evidence versions are immutable; corrections create a new version so the decision trail remains auditable.

The portfolio and detail page highlight evidence gaps such as missing ownership, Work, evidence, or an outcome value/explicit limitation. Clear these warnings before using the assessment in a client decision.

Compare scenarios

Open Capabilities from inside a scenario to see the scenario state and its movement from the comparison source. Scenario links stay in context automatically. Definitions remain shared across the tenant, while assessments, Work mappings, measure values, and eligible evidence links belong to their Live or scenario context.

Snapshots and archived scenarios are read-only. They retain the assessment and proof as an auditable record; editing controls are unavailable.

Use capabilities across agency clients

Create and assess capabilities inside the client's own tenant. An agency's consultants can work across several client tenants, but the capability portfolio is not shared between them. This isolation prevents one client's definitions, evidence, owners, or performance measures from appearing in another client's workspace.

There is currently no agency-wide capability library or cross-tenant clone. Recreate a suitable high-level capability framework in each client tenant, then tailor the outcomes, language, measures, and assessments to that client. Never copy client evidence, measured values, ownership, or confidential rationale into another tenant. Scenario cloning remains the fast way to reuse and evolve the model within one client tenant.

Archive and restore

Archive capabilities that are no longer part of the active portfolio. History and contextual records are retained. Move or archive active children first. Restore the parent before restoring one of its children.

Read capabilities through API, Astro, MCP, or CLI

The read API uses the same scenario-aware semantic projection as the web workspace:

• GET /api/v1/capabilities lists capability states and deltas. Use scenario=<slug> for a target scenario, page/per_page for pagination, include for selected top-level sections, and fields[capabilities] for sparse top-level fields.
• GET /api/v1/capabilities/{slug} returns one capability with its state, source delta, measures, mapped Work, matched deployments, evidence, and completeness findings.

Metrics and costs remain independently permission-filtered. Astro uses read.business-capability.list to browse or search the catalog in Live, a scenario, or a snapshot, then read.business-capability.detail to inspect one result. It can also resolve a capability by name or use the capability page currently open as its focus. MCP exposes the governed detail read; neither assistant surface can write a capability. In the Orgonaut CLI, use orgonaut capability list or orgonaut capability show <slug>. A saved scenario profile automatically adds the scenario slug, while Live mode remains explicitly baseline-scoped.

Troubleshooting

• Capabilities is not in the navigation: confirm your plan, rollout access, and capabilities.read permission.
• Editing controls are missing: the context may be a snapshot, archived scenario, or your role may be read-only.
• No ownership choices are available: add a team or department for organisational ownership, or add a position for human accountability, in the same Live or scenario context as the assessment. Adding a person alone does not create an accountable position.
• Work cannot be mapped: save the contextual assessment first and make sure the Work activity exists in the same context.
• Evidence cannot be attached: save the assessment first and use evidence that is allowed in the active context.
• A measure value is rejected: provide either one point or one complete range, use the correct canonical unit, and do not date observed evidence in the future.

Related guides: Transformation Cases, AI solutions and deployments, Work catalog, Create a scenario, and Cost and velocity deltas.