Field Service Foundation Setup Designer
Design or refine Field Service foundation data — Work Types, Skills, Service Territories, Operating Hours, and other associated Field Service objects — through structured interviews. Starting from scratch or from existing recommendations.
This skill runs a design conversation — it does not call Salesforce APIs directly. It collects the design through a structured conversation, whether starting from scratch or working from existing data, and confirms the final design with the user. This skill covers two design workflows in sequence (or independently):
- Work Types & Skills — 6-question interview
- Service Territories & Operating Hours — 3-question interview
Entry point
- Starting from scratch → run the full interview phases below to build the design.
- Have existing data — this includes scraped/read source output, JSON or text provided by the user, or any other external source brought into context. Show a plain-English summary of the proposed configuration as a starting reference, then immediately ask: "Would you like to go through a guided design interview to refine this, or proceed directly with this configuration?" Wait for the answer before doing anything else.
- Ready to create → show a plain-English summary of the records to be created (not raw JSON or skill names), ask for explicit confirmation, then proceed only after the user confirms.
Never expose internal skill names, SOR IDs, or tool references to the user.
When to use
- Starting Field Service setup from scratch — no existing Work Types or Territories yet
- When initial recommendations need adjustment for either work types or territories (or both)
- To explore different granularity levels (Work Types) or territory models (Territories)
- When customer provides additional context about their business or coverage areas
- During discovery sessions to design or iterate on the foundation data model
Design principles
Structured questions with consistent intent
- Questions based on Work Type Implementation Guide and Service Territory Design Guide
- Adapt examples to customer's business context
- Intent stays consistent, wording adapts intelligently
Brief answers must work
- "No", "Yes", "No we don't" are complete answers
- Progress to next question immediately
- No explanations required
Smart question adaptation
- Check if topic already covered
- If covered: brief confirmatory question
- If not covered + business context: customize examples
- If not covered + no context: exact structured question
Question format (three or four parts):
**Recommendation:** [Only include when context is sufficient. Lead with the best-fit option given what's known, then briefly note alternatives with the condition under which they'd apply instead — not a flat menu of equal choices, but a ranked steer. Omit entirely when confidence is low.]
**Question:** [The actual question — or a lighter "does this fit?" when a Recommendation is present]
**Why I'm asking:** [1-2 sentences: why this matters]
**Impact:** [1-2 sentences: how their answer affects the design]
How it works
Pre-phase: Source ingestion (if provided)
If no source has been provided, creatively ask the user whether they have a website or document that describes their business or Field Service operation — framing it as something that would help build a more accurate and tailored starting recommendation rather than a generic one.
If the user provides a URL, fetch it. Extract Field Service-relevant signals: industry, service types, geographic coverage, team and skill structure. Use these to generate an initial recommendation covering Work Types, Skills, Service Territories, and Operating Hours — which feeds into the refinement interview rather than starting from nothing.
Phase 1: Work Types & Skills
- Accept input: Takes existing data in any format (JSON, CSV, free-form text), or starts fresh if nothing is provided.
- Interview: Cover these topics — service line granularity, equipment size/type variations, brand/model specificity, service tiers and site types, parts tracking, duration accuracy.
- Output: Designed Work Types and Skills as JSON + change log.
Phase 2: Service Territories & Operating Hours
- Accept input: Same as Phase 1.
- Interview: Cover these topics — territory structure (geographic/functional/hybrid), geographic boundaries (if applicable), functional team geographic constraints (if applicable).
- Output: Designed Service Territories and Operating Hours as JSON + change log.
Inputs
- Source material (optional): A URL the agent will fetch to extract business context and generate an initial recommendation. Takes precedence over manually pasted data when both are provided.
- Existing data (optional): Work Types/Skills and/or Service Territories/Operating Hours in any format — JSON, CSV, free-form text, or pasted output from a prior session. If nothing is provided, recommendations are built entirely through the interview.
- Business context (optional): Additional context to tailor interview questions to the customer's industry or service model.
Outputs
- Designed or refined Work Types with metadata
- Skills inventory organized by category
- Designed or refined Service Territories with structure
- Operating Hours organized by timezone
- Combined JSON export (Salesforce API-ready, for deployment via
sfs-sobject-create)
- Markdown summary with change log
- Before/after comparison (when refining existing data)
Granularity Levels Reference (Work Types)
From Work Type Implementation Guide:
| Level |
When to Use |
Example |
| High-Level |
Simple service models, quick setup |
"Installation", "Repair", "Maintenance" |
| Mid-Level |
Equipment-specific service |
"HVAC Installation", "HVAC Repair" |
| Detailed |
Brand-specific service needs |
"HVAC Installation (Carrier)", "HVAC Installation (Trane)" |
| Hyper-Specific |
SLA/site variations |
"HVAC Repair (Carrier) - Hospital" |
Best practice: Start with simplest model possible, increase granularity only when necessary.
Territory Models Reference (Service Territories)
From Service Territory Design Guide:
| Model |
When to Use |
Example |
| Geographic |
Location-based coverage, reduce travel time |
"Northern California", "Southwest Region" |
| Functional |
Specialized service types, skill-based assignment |
"Fire Safety Team", "Commercial HVAC Team" |
| Hybrid |
Combined benefits, specialized teams with regional boundaries |
"Northern CA - Fire Safety", "Southwest - Commercial HVAC" |
Best practice: Start with simplest model possible, increase complexity only when necessary.
Output format
Generate the output JSON payload from interview answers using this mapping:
WorkType → SkillRequirement (junction → Skill)
WorkType → ProductRequired (junction → Product2)
OperatingHours → TimeSlot (parent-child)
ServiceTerritory → OperatingHours (lookup)
1---2name: field-service-foundation-setup-designer-get3description: Creates Field Service foundation data — Work Types, Skills, Service Territories, and Operating Hours. Accepts existing data in any format or designs from scratch, and confirms the design with the user. Use this skill when a user wants to create, design, or set up Field Service foundation data.4---5
6# Field Service Foundation Setup Designer
7
8**Design or refine Field Service foundation data — Work Types, Skills, Service Territories, Operating Hours, and other associated Field Service objects — through structured interviews. Starting from scratch or from existing recommendations.**
9
10This skill runs a design conversation — it does not call Salesforce APIs directly. It collects the design through a structured conversation, whether starting from scratch or working from existing data, and confirms the final design with the user. This skill covers two design workflows in sequence (or independently):
11
121. **Work Types & Skills** — 6-question interview
132. **Service Territories & Operating Hours** — 3-question interview
14
15## Entry point
16
17- **Starting from scratch** → run the full interview phases below to build the design.
18- **Have existing data** — this includes scraped/read source output, JSON or text provided by the user, or any other external source brought into context. Show a plain-English summary of the proposed configuration as a starting reference, then immediately ask: "Would you like to go through a guided design interview to refine this, or proceed directly with this configuration?" Wait for the answer before doing anything else.
19- **Ready to create** → show a plain-English summary of the records to be created (not raw JSON or skill names), ask for explicit confirmation, then proceed only after the user confirms.
20
21Never expose internal skill names, SOR IDs, or tool references to the user.
22
23## When to use
24
25- Starting Field Service setup from scratch — no existing Work Types or Territories yet
26- When initial recommendations need adjustment for either work types or territories (or both)
27- To explore different granularity levels (Work Types) or territory models (Territories)
28- When customer provides additional context about their business or coverage areas
29- During discovery sessions to design or iterate on the foundation data model
30
31## Design principles
32
331. **Structured questions with consistent intent**
34 - Questions based on Work Type Implementation Guide and Service Territory Design Guide
35 - Adapt examples to customer's business context
36 - Intent stays consistent, wording adapts intelligently
37
382. **Brief answers must work**
39 - "No", "Yes", "No we don't" are complete answers
40 - Progress to next question immediately
41 - No explanations required
42
433. **Smart question adaptation**
44 - Check if topic already covered
45 - If covered: brief confirmatory question
46 - If not covered + business context: customize examples
47 - If not covered + no context: exact structured question
48
494. **Question format** (three or four parts):
50 ```text
51 **Recommendation:** [Only include when context is sufficient. Lead with the best-fit option given what's known, then briefly note alternatives with the condition under which they'd apply instead — not a flat menu of equal choices, but a ranked steer. Omit entirely when confidence is low.]
52
53 **Question:** [The actual question — or a lighter "does this fit?" when a Recommendation is present]
54
55 **Why I'm asking:** [1-2 sentences: why this matters]
56
57 **Impact:** [1-2 sentences: how their answer affects the design]
58 ```
59
60## How it works
61
62### Pre-phase: Source ingestion (if provided)
63
64If no source has been provided, creatively ask the user whether they have a website or document that describes their business or Field Service operation — framing it as something that would help build a more accurate and tailored starting recommendation rather than a generic one.
65
66If the user provides a URL, fetch it. Extract Field Service-relevant signals: industry, service types, geographic coverage, team and skill structure. Use these to generate an initial recommendation covering Work Types, Skills, Service Territories, and Operating Hours — which feeds into the refinement interview rather than starting from nothing.
67
68### Phase 1: Work Types & Skills
69
701. **Accept input**: Takes existing data in any format (JSON, CSV, free-form text), or starts fresh if nothing is provided.
712. **Interview**: Cover these topics — service line granularity, equipment size/type variations, brand/model specificity, service tiers and site types, parts tracking, duration accuracy.
723. **Output**: Designed Work Types and Skills as JSON + change log.
73
74### Phase 2: Service Territories & Operating Hours
75
761. **Accept input**: Same as Phase 1.
772. **Interview**: Cover these topics — territory structure (geographic/functional/hybrid), geographic boundaries (if applicable), functional team geographic constraints (if applicable).
783. **Output**: Designed Service Territories and Operating Hours as JSON + change log.
79
80
81## Inputs
82
83- **Source material** (optional): A URL the agent will fetch to extract business context and generate an initial recommendation. Takes precedence over manually pasted data when both are provided.
84- **Existing data** (optional): Work Types/Skills and/or Service Territories/Operating Hours in any format — JSON, CSV, free-form text, or pasted output from a prior session. If nothing is provided, recommendations are built entirely through the interview.
85- **Business context** (optional): Additional context to tailor interview questions to the customer's industry or service model.
86
87## Outputs
88
89- Designed or refined Work Types with metadata
90- Skills inventory organized by category
91- Designed or refined Service Territories with structure
92- Operating Hours organized by timezone
93- Combined JSON export (Salesforce API-ready, for deployment via `sfs-sobject-create`)
94- Markdown summary with change log
95- Before/after comparison (when refining existing data)
96
97## Granularity Levels Reference (Work Types)
98
99From Work Type Implementation Guide:
100
101| Level | When to Use | Example |
102|-------|-------------|---------|
103| **High-Level** | Simple service models, quick setup | "Installation", "Repair", "Maintenance" |
104| **Mid-Level** | Equipment-specific service | "HVAC Installation", "HVAC Repair" |
105| **Detailed** | Brand-specific service needs | "HVAC Installation (Carrier)", "HVAC Installation (Trane)" |
106| **Hyper-Specific** | SLA/site variations | "HVAC Repair (Carrier) - Hospital" |
107
108**Best practice:** Start with simplest model possible, increase granularity only when necessary.
109
110## Territory Models Reference (Service Territories)
111
112From Service Territory Design Guide:
113
114| Model | When to Use | Example |
115|-------|-------------|---------|
116| **Geographic** | Location-based coverage, reduce travel time | "Northern California", "Southwest Region" |
117| **Functional** | Specialized service types, skill-based assignment | "Fire Safety Team", "Commercial HVAC Team" |
118| **Hybrid** | Combined benefits, specialized teams with regional boundaries | "Northern CA - Fire Safety", "Southwest - Commercial HVAC" |
119
120**Best practice:** Start with simplest model possible, increase complexity only when necessary.
121
122---
123
124## Output format
125
126Generate the output JSON payload from interview answers using this mapping:
127
128- `WorkType` → `SkillRequirement` (junction → `Skill`)
129- `WorkType` → `ProductRequired` (junction → `Product2`)
130- `OperatingHours` → `TimeSlot` (parent-child)
131- `ServiceTerritory` → `OperatingHours` (lookup)