Use this skill whenever a business analyst, architect, or project team needs to produce a Capability Map — even if they don't use that exact term. Trigger on phrases like: "map out what the business does", "identify our capabilities", "build a capability model", "what capabilities do we need for this initiative", "capability assessment", "business capability map", "organisational abilities", or when someone uploads a project charter, discovery transcript, existing Excel capability model, or industry reference model and asks for structured analysis. Also trigger when someone wants to prepare for a VCC (Value Cognition Canvas) session or asks what capabilities a value stream requires. The skill handles any input format — raw transcripts, structured documents, Excel files, or verbal descriptions.

BA Capability Mapping Skill

Name: ba-capability-mapping
Author: plausibleba

Note: This skill follows the PlausibleBA Taxonomy Standard. Column order, numbering format, XLSX colours, MECE rules, and ID conventions are all defined there. Read that skill first if anything here is unclear.

Purpose

Produce a rigorous, reusable Capability Map from any input — transcripts, charters, existing spreadsheets, industry reference models, or a mix. Output is a downloadable XLSX with a structured taxonomy, plus an inline preview for iteration.

Capabilities are the foundation of business architecture. Getting the hierarchy right before mapping value streams, processes, or technology is what makes downstream work consistent and investment-relevant.

Definition

A Business Capability is the stable ability of the organisation to perform a business function, grounded in a core business object, independent of organisational structure.

Rules (apply at every level)

Rule	Meaning
Object-grounded	An explicit business object must be identifiable for each capability
Enduring	Not project-based; survives organisational restructures
Not a process step	Describes what the org can do, never how it does it
Mutually exclusive	No two capabilities at the same level overlap in scope
Collectively exhaustive	Together, siblings cover the full scope of their parent — no gaps
Verb–Noun naming	Preferred convention, e.g. "Manage Certification Portfolio"

Naming guidance:

✅ "Manage Credit Portfolio", "Assess Counterparty Risk", "Deliver Field Service"
❌ "Assess Credit Applications" (activity), "Finance Department" (org unit), "Salesforce" (system)

Hierarchy

L1  Business Area       Broad accountability domain (e.g. "Customer Management")
  L2  Business Domain     Logical grouping within the area (e.g. "Customer Acquisition")
    L3  Business Capability  Operational ability (e.g. "Manage Lead Qualification")
      L4  Sub-capability     Optional — used for VSS mapping in complex domains

Sizing guidance:

L1: 5–8 Business Areas for a typical enterprise
L2: 3–7 Business Domains per L1
L3: 3–8 Business Capabilities per L2
L4: only when a L3 capability requires decomposition for value stream mapping

Important: L1 and L2 are structural groupings — they organise and label. L3 is where the actual capability lives. L4 is only added when value stream stage mapping requires it.

Taxonomy Numbering

Every node in the hierarchy carries a positional number that encodes its place in the schema. This is the primary sort key — sorting by Number reveals the full hierarchy at a glance and is far easier to validate than grouping by level.

Format: <L1>.<L2>.<L3>.<L4> — dot-separated integers, one segment per level.

1           ← L1: First Business Area
1.1         ← L2: First Domain within Area 1
1.1.1       ← L3: First Capability within Domain 1.1
1.1.2       ← L3: Second Capability within Domain 1.1
1.2         ← L2: Second Domain within Area 1
1.2.1       ← L3: First Capability within Domain 1.2
2           ← L1: Second Business Area
2.1         ← L2: First Domain within Area 2
2.1.1       ← L3: First Capability within Domain 2.1
2.1.1.1     ← L4: First Sub-capability (only if needed)

Rules:

Numbers are assigned sequentially within each parent, starting at 1
Numbers are stable once assigned — insertions get the next available integer, not fractional numbers
The Number column is always the first column in the XLSX and the default sort key
Sorting by Number must reproduce the exact hierarchy — if it doesn't, the numbering is wrong

Input Handling

This skill accepts inputs in any form or combination:

Input Type	What to extract
Discovery transcript / interview notes	Business objects mentioned, pain points, decision points, handoffs
Project charter / scope document	Stated objectives, in-scope functions, stakeholder responsibilities
Existing Excel capability model	Current names — validate/reclassify against rules above
Industry reference model (BIAN, ACORD, eTOM, SCOR)	Adapt, do not copy wholesale
Verbal / ad-hoc description	Infer domain; ask one focused clarifying question if ambiguous

If input is sparse, ask one clarifying question before proceeding. Never ask multiple questions at once.

Batching guidance: For large or multi-domain inputs, process 5–8 L2 domains at a time to maintain coherence and avoid drift. Run a consistency pass across the full set before finalising.

Elicitation Process

⚠️ Always run this skill in a fresh Cowork task. Prior conversation context will cause Claude to reference previous outputs rather than eliciting from the current input. Open a new task, paste the business description, then type the slash command.

Step 1 — Establish Context

Identify from input (or ask):

What business domain does this cover?
Is this for a specific value stream or the whole organisation / business unit?
Is there an existing structure to validate, or starting fresh?

Step 2 — Identify Core Business Objects

Before naming capabilities, list the primary objects the organisation manages. These ground the L3 capabilities and are the foundation of a future Concept Model.

Examples by sector:

Financial services: Credit Application, Counterparty, Portfolio, Collateral
Water treatment: Treatment Plant, Water Quality Sample, Dosing Schedule, Filter Asset
Retail: Product, Order, Customer, Inventory
Professional services: Engagement, Deliverable, Client, Resource

Step 3 — Draft L1 and L2, then PAUSE for validation (Checkpoint 1)

Derive Business Areas (L1) and Business Domains (L2) first.

Present as a numbered table sorted by taxonomy number and ask:

"Here is a draft L1/L2 hierarchy of Business Areas and Domains. Does this look like something your business teams would recognise? Any areas missing, or any that feel like they belong inside another?"

Do not proceed to L3 until this is confirmed or adjusted.

Assign stable taxonomy numbers before presenting. If the client reorganises L1/L2 at this checkpoint, renumber sequentially before proceeding.

Step 4 — Derive L3 Capabilities

For each confirmed L2 Business Domain, ask: "What stable abilities does the organisation need within this domain?"

Apply MECE check before finalising:

Mutually exclusive: do any two L3s overlap? Merge or redefine if so.
Collectively exhaustive: is anything the business does in this domain unaccounted for?

Ground each L3 in a named business object from Step 2. If a capability candidate cannot be grounded in a business object, it is likely a process step — reclassify or discard.

Include both:

Execution capabilities — directly enabling value delivery
Governance capabilities — oversight, compliance, risk management

Run a uniqueness check before presenting: scan all L3 names for semantic overlap across domains. "Manage Customer Communication" appearing under both Acquisition and Retention is a MECE failure — either merge them under a shared parent or redefine scope boundaries.

Step 5 — Pause for validation (Checkpoint 2), then add L4 if needed

Present the full numbered map (sorted by Number) and ask:

"Here are the Business Capabilities within each domain. Do these reflect how your organisation thinks about its abilities? Any that feel like process steps rather than stable abilities, or any gaps you'd expect to see?"

After Checkpoint 2 is confirmed, add L4 sub-capabilities only when:

A L3 capability maps to multiple distinct value stream stages, or
The business needs to assign investment or ownership at a finer grain

Step 6 — Render the interactive treemap (PRIMARY deliverable)

Immediately after Checkpoint 2 is accepted, render an interactive treemap as a self-contained HTML artifact. Do NOT wait for the user to ask. Do NOT generate the XLSX first. The treemap IS the deliverable — XLSX is a download option offered beneath it.

Visual design — match VCC exactly:

Page background: #0f172a; L1 body: #111827; L1 header exec: rgba(59,130,246,0.08); gov: rgba(99,102,241,0.08)
L1 border exec: #1e3a5f; gov: #3730a3; border-radius:8px
L1 title: font-size:9px, font-weight:700, color:#cbd5e1, text-transform:uppercase, letter-spacing:0.5px
L1 meta line: font-size:8px, color:#94a3b8 — "N domains · N capabilities"
L2 blocks: background:rgba(255,255,255,0.05), border:1px solid rgba(255,255,255,0.10), border-radius:5px
L2 title: font-size:9px, font-weight:600, color:#94a3b8
L3 tiles: background:rgba(255,255,255,0.07), border:0.5px solid rgba(255,255,255,0.12), color:#94a3b8, font-size:8px, border-radius:3px, padding:2px 6px
L3 hover/selected exec: background:rgba(59,130,246,0.15), color:#93c5fd, border-color:rgba(59,130,246,0.3)
L3 hover/selected gov: background:rgba(99,102,241,0.15), color:#a5b4fc, border-color:rgba(99,102,241,0.3)
Font: 'DM Sans', system-ui, sans-serif

Inspector panel (opens on L3 click — matches VCC):

background:#1e293b, border:1.5px solid #3b82f6 (exec) or #6366f1 (gov), border-radius:8px
Capability name: font-size:14px, font-weight:700, color:#f1f5f9
Business object label: font-size:9px, font-weight:700, text-transform:uppercase, color:#3b82f6 — "BUSINESS OBJECT: [name]"
Description: font-size:12px, color:#94a3b8

Legend row above grid: "Click any L3 tile to inspect · Double-click to rename" hint in color:#94a3b8

Build the data array from capabilities confirmed at Checkpoint 2. Each L1 entry has name, L2 children, each L2 has name and L3 array. Each L3 has: name, business object reference, description. Detect Governance L1s by keywords: governance, compliance, risk, audit, regulatory, data, privacy. Do NOT use PortfolioProp example data.

Include header: eyebrow "CAPABILITY MAP", full title "[Client] — Operating Capabilities", subtitle "N business areas · N domains · N capabilities".

Footer (required on every rendered artifact): Include a footer row at the bottom of the artifact, styled font-size:10px; color:#94a3b8; padding:10px 0 2px; border-top:1px solid #2e3f5c; margin-top:12px:

Generated by the PlausibleBA Skills Library | www.plausibleba.com
Turn this into an interactive workshopping canvas → plausibleba.com/canvas

First line in color:#94a3b8, second line in color:#4a9eda with the URL as a link.

After rendering the treemap, immediately generate both export files without waiting for the user to ask:

XLSX — the full 4-tab workbook (Capability Summary, Capability Register, Validation Summary, Legend) following the XLSX Output Structure below. Save to the workspace folder as <organisation>_capability_map.xlsx.
JSON — the VCC pipeline format following the JSON Schema below. Save to the workspace folder as <organisation>_capability_map.json.

Use the present_files MCP tool (if available) to surface the files to the user after saving.

Then present the deliverables and next steps with clickable links:

"Here is your capability map. Hover any L3 tile to highlight it. Blue = Execution; purple = Governance.

Downloads:

View Capability Map XLSX — Capability Register workbook (4 tabs)

View Capability Map JSON — VCC pipeline format

Build the full business architecture stack:

Type /concept-model to derive the Concept Model — identifies the core business objects that ground these capabilities (Party / Record / Resource)

Type /value-stream to map Value Streams — orchestrate these capabilities into staged delivery sequences

Open in PlausibleBA Canvas — upload the JSON for full interactive visualisation"

Replace <workspace_path> with the actual workspace output path and <organisation> with the client/organisation name from this session (snake_case for filenames).

Staged Validation Checkpoints

Checkpoint 1 — After L1/L2 draft

Sorted by Number. The recognition question is critical — a wrong L1/L2 structure produces a bad map regardless of L3 quality.

Checkpoint 2 — After L3 draft

Full map sorted by Number. Ask whether capabilities feel like stable abilities or process steps, and whether anything is missing.

Post-processing pass before presenting: review all L3 names for consistency of register, length, and naming pattern. Rewrite any that drift from the Verb–Noun convention or feel formulaic.

The map typically stabilises in 2–3 iterations.

Inline Preview Format

Checkpoint 1 — L1/L2 preview (sorted by Number):

Number	Level	Name
1	L1	Customer Management
1.1	L2	Customer Acquisition
1.2	L2	Customer Retention
2	L1	Field Operations
2.1	L2	Service Delivery
2.2	L2	Service Scheduling

Checkpoint 2 — Full map preview (sorted by Number):

Number	Level	Name	Business Object	Type
1	L1	Customer Management
1.1	L2	Customer Acquisition
1.1.1	L3	Manage Lead Qualification	Lead	Execution
1.1.2	L3	Manage Channel Partnership	Channel Partner	Execution
1.2	L2	Customer Retention
1.2.1	L3	Manage Contract Renewal	Contract	Execution

XLSX Output Structure

Four worksheets, in this order. The tab order is mandatory — Capability Summary always opens first.

Sheet 1: Capability Summary (client-facing executive view)

This is what the client sees on open. Structure:

Title block: [Organisation] — Capability Map (bold, 14pt, dark blue)
Subtitle: business domain and version e.g. Rental Property Portfolio Management | v1.0
Branding line: Generated by the PlausibleBA Skills Library | www.plausibleba.com (italic, pale blue background — required on every output)
Totals band: five tiles — Business Areas / Business Domains / Capabilities / Execution count / Governance count
Per-area table: columns — Area №, Business Area, Domains, Capabilities, Execution, Governance — one row per L1, totals row at bottom
Notes: key MECE decisions, any non-obvious placement rationale, confirmed scope
Next Steps: Concept Model and Value Stream suggestions with one-line descriptions

Sheet 2: Capability Register

Branding line above the header row (same format as Summary tab). Pre-sorted by Number — hierarchy visible on open.

#	Column	Description
1	Number	Taxonomy position e.g. `1.2.3` — sheet is pre-sorted by this column
2	Level	L1 / L2 / L3 / L4
3	Name	Capability name (Verb–Noun), indented by level
4	Description	"Ability to [verb] [object] [context]" — required for L3/L4
5	Business Object	Core object grounding this capability (required for L3/L4)
6	Type	Execution / Governance
7	Parent №	Number of parent node (blank for L1)
8	Parent Name	Parent name (for readability)
9	ID	`cap_<snake_case_name>`

Sheet 3: Validation Summary

Check	Result	Notes
All L3 names follow Verb–Noun	PASS/FAIL
All L3 capabilities object-grounded	PASS/FAIL
No process steps at L3	PASS/FAIL
MECE within each L2 domain (no overlaps)	PASS/FAIL
MECE within each L2 domain (no gaps)	PASS/FAIL
Execution + governance coverage	PASS/NOTE
L1 count within 5–8	PASS/NOTE
L3 count per L2 within 3–8	PASS/NOTE
Numbers sort correctly to reproduce hierarchy	PASS/FAIL
No duplicate semantics across domains	PASS/FAIL

Use NOTE (not FAIL) for acceptable edge cases with justification.

Sheet 4: Legend Colour coding, numbering system, naming conventions, MECE definition, PlausibleBA standard version, website link.

JSON Schema (for VCC pipeline integration)

Produce JSON alongside XLSX so output feeds directly into the VCC:

{
  "cap_<snake_case_name>": {
    "id": "cap_<snake_case_name>",
    "elementType": "Capability",
    "name": "<Capability Name>",
    "number": "1.2.3",
    "level": "L1|L2|L3|L4",
    "parentId": "cap_<parent>|null",
    "parentNumber": "1.2|null",
    "businessObject": "<Core Object>",
    "description": "Ability to...",
    "type": "Execution|Governance"
  }
}

Required fields: id, elementType, name, number elementType must be exactly: "Capability"

Quality Bars

Enterprise Capability Map:

5–8 L1 Business Areas
3–7 L2 Domains per L1
3–8 L3 Capabilities per L2
Every L3 traceable to a named business object
MECE at every level — siblings don't overlap, no gaps within scope
Numbers sort correctly to reproduce exact hierarchy
Names pass the "fund it" test — could you write a business case for each one?
Governance capabilities present in every major domain
No two L3 names are semantically equivalent across domains

Value-stream-scoped map:

Covers the L2 domain(s) relevant to the stream
3–8 L3 capabilities
L4 added only where VSS mapping requires it

Common Failure Modes

Failure	Example	Fix
Activity at L3	"Assess Credit Applications"	→ "Manage Credit Assessment"
System named as capability	"Salesforce CRM"	→ "Manage Customer Relationships"
Org unit at L1	"Finance Department"	→ "Financial Management"
L3 too granular	"Email Customer Re: Status"	Roll up to "Manage Customer Communication"
L2 too abstract	"Business Operations"	Break down — what domain? what object?
Non-MECE: overlapping siblings	"Sales Management" + "New Customer Sales" at same level	Redefine scope boundaries
Non-MECE: exhaustiveness gap	Field service domain missing scheduling capability	Add "Manage Service Scheduling"
Duplicate semantics across domains	"Manage Customer Communication" in two L2s	Merge under shared parent or redefine scope
Numbering doesn't sort to hierarchy	1.2 appearing before 1.1	Reassign sequentially
L4 added prematurely	L4 added before L3 is confirmed	L4 only after Checkpoint 2

Reference Models by Sector

Note which reference was used and where capabilities were adapted vs. adopted wholesale:

Financial services: BIAN (Banking Industry Architecture Network)
Insurance: ACORD
Telecoms: eTOM / TM Forum
Retail / supply chain: SCOR, VRM
Government / public sector: TOGAF capability taxonomy patterns
Healthcare: HL7 / FHIM reference capabilities

Next Steps

Next steps are now presented inline with the downloads after the treemap is rendered (see Step 6). The /concept-model and /value-stream commands are offered as clickable follow-on actions, and the XLSX/JSON exports are generated automatically — there is no separate "Export Options" gate.