Process Mapping

Generate structured process map diagrams from process descriptions. Each step is a block with action, actor, and tool — visually separated within the block.

This skill draws diagrams. It does NOT analyze processes, identify bottlenecks, recommend automation, or estimate ROI — that's the business-consultant's job. If the user asks for process analysis, direct them to the business-consultant skill and offer to generate diagrams from its output.

Output Formats

1. Excalidraw file (default) — .excalidraw JSON file, opens in Obsidian (Excalidraw plugin), VS Code, excalidraw.com
2. Mermaid — when user explicitly requests it. Output inline in response (do NOT generate separate .mmd files — Mermaid renders directly in chat)

Default to Excalidraw unless the user says otherwise.

Context Dependencies

If context/process-mapping.md is missing, use defaults and note: "Brakuje pliku context/process-mapping.md (preferencje kolorów, format output, ścieżki). Używam ustawień domyślnych. Uruchom /environment-setup aby skonfigurować."

Workflow

Step 1: Understand the Input

The user provides process steps — either:

• Structured — numbered list with actors and tools (ready to diagram)
• Unstructured — meeting notes, narrative, or transcript

For unstructured input, extract steps into a structured list and present for confirmation. Mark any information that is inferred or assumed (not explicitly stated) with [assumed] in the text response, and with ⚠️ on the diagram:

Extracted process (5 steps):
1. [Receive complaint] Actor: BOK Employee | Tool: Phone/Email
2. [Log ticket] Actor: BOK Employee | Tool: Ticketing System [assumed]
3. [Check order history] Actor: BOK Employee | Tool: CRM
4. [Approve?] Decision → Yes: step 5, No: step 6
5. [Generate refund] Actor: BOK Employee | Tool: ERP
6. [Send rejection] Actor: BOK Employee | Tool: Email [assumed]

Important: Do NOT add business analysis, bottleneck identification, or automation recommendations. If the user wants that, suggest using the business-consultant skill first, then come back here for diagrams.

Step 2: Load Color Palette

Check for colors in this order (see references/color-palette.md):

1. Project design system (design-system.md or design-tokens.css)
2. Context overrides (context/process-mapping.md)
3. Default semantic palette

Step 3: Generate Diagram

Excalidraw Output (Default)

Generate .excalidraw JSON file following:

• references/json-schema.md — file structure, element properties
• references/element-templates.md — block templates, decision diamonds
• references/arrows-and-layout.md — layout patterns, arrow routing
• references/color-palette.md — semantic colors

Orientation: LEFT-TO-RIGHT (horizontal flow). This is the default. Use vertical flow only if the user requests it or if the process is very short (≤4 steps).

Block format — every process step is a rectangle with three visually separated sections:

┌─────────────────────┐
│  Receive Complaint   │  ← action (bold, larger font)
├─────────────────────┤
│  Actor: BOK Employee │  ← actor (smaller, muted color)
│  Tool: Email         │  ← tool (smaller, muted color)
└─────────────────────┘

The action name is visually dominant. Actor and tool are secondary information below a divider line.

Decision blocks — use rotated squares (45° diamond shape). Since raw Excalidraw JSON breaks arrow bindings on diamonds, implement as a rectangle with angle: 0.785 (π/4 radians) rotation. Keep text horizontal inside.

Color coding for step types:

• Manual step — blue palette (default)
• Automated/system step — purple palette
• Problem/bottleneck step — red palette (only if explicitly flagged by user or another skill)

Single file policy: When generating multiple related diagrams (AS-IS + TO-BE), put them ALL on one canvas in the same .excalidraw file. Position them side-by-side with clear section titles. Do not create separate files.

Save to project directory. Default: docs/processes/[process-name].excalidraw

Build section by section for large diagrams (>8 steps).

Mermaid Output

When explicitly requested, output Mermaid flowchart inline in your response (in a ```mermaid code block). Do NOT create separate .mmd files.

flowchart LR
    START([Start: Complaint received])
    S1["Receive Complaint\nActor: BOK Employee\nTool: Email"]
    D1{"Complaint valid?"}
    S2["Generate Refund\nActor: BOK Employee\nTool: ERP"]
    S3["Send Rejection\nActor: BOK Employee\nTool: Email"]
    END1([End: Refund issued])
    END2([End: Rejection sent])

    START --> S1 --> D1
    D1 -->|Yes| S2 --> END1
    D1 -->|No| S3 --> END2

Note: use flowchart LR (left-to-right) as default direction.

Step 4: Validate

Run through references/validation.md checklist. Report only if issues found — don't output validation details when everything passes.

Step 5: AS-IS / TO-BE (if requested)

When the user provides both current and target process:

1. Draw both on one canvas — AS-IS on the left, TO-BE on the right
2. Use section titles: "AS-IS: [name]" and "TO-BE: [name]"
3. Visual differentiation:
   • Changed steps → purple (automated)
   • Removed steps → dashed red border
   • Unchanged steps → same as AS-IS
4. Do NOT add automation analysis — just visualize what the user described

Response Format

Keep the response concise:

• Show extracted steps (if from unstructured input)
• State where the file was saved
• Note any assumptions marked with [assumed]

Do NOT include:

• Color legends or palette descriptions
• "Output files" sections listing what was generated
• Validation details (unless something failed)
• Business analysis, bottleneck analysis, or ROI estimates

Boundaries

• This skill draws process diagrams — it does not analyze or optimize processes
• For process analysis → use business-consultant skill
• NOT for: architecture diagrams, ER diagrams, sequence diagrams, mind maps, org charts
• Process complexity: optimized for 3-20 steps. For 20+ steps, suggest sub-processes
• Language: supports Polish and English

Commands

• /process or /mapa — invoke this skill
• /mermaid — force Mermaid output instead of Excalidraw

Routing text before the description was cut

The description in the front matter is capped at 250 characters, because the list of skills a model sees has a character budget. The full routing text it carried before — triggers, boundaries, what the skill is not for — is kept here:

Draw process flow diagrams showing who does what with which tool — generates Excalidraw (.excalidraw) or Mermaid diagrams from process descriptions. ONLY for sequential business process flows (step-by-step workflows with actors and tools). Use when the user wants to: map a business process, draw a process flowchart, create AS-IS or TO-BE process maps, visualize a workflow as a diagram, document process steps from meeting notes or transcripts, generate a process map showing actors and tools per step. Triggers: 'zmapuj proces', 'mapa procesu', 'diagram przepływu', 'narysuj flowchart procesu', 'process map', 'workflow diagram', 'AS-IS TO-BE', 'mapowanie procesu', 'kroki procesu', 'flowchart', 'narysuj diagram procesu', 'udokumentuj proces', 'process flow', 'zmapuj workflow'. NOT for: architecture diagrams, sequence diagrams, ER diagrams, org charts, mind maps, or any non-process diagrams — even if the user asks for excalidraw output. NOT for process analysis, bottleneck identification, or automation recommendations — that is the business-consultant skill's job. This skill draws diagrams only. Also trigger when business-consultant or another skill needs a process diagram drawn.