tomevault-io City Policy Analysis — Master Orchestrator
Four-Phase Integrated Framework
| Phase | Source Methodology | When to Use | Reference File |
|---|---|---|---|
| 1. FRAME | Bloomberg Center for Public Innovation (JHU) | Problem is undefined or needs scoping | Problem_Framing_Skill.md |
| 2. ANALYZE | J-PAL, MIT — Evidence-to-Policy | Running numbers, finding patterns, equity analysis | Analytical_Skill.md |
| 3. COMMUNICATE | The GovLab (NYU) / InnovateUS | Writing memos, briefs, dashboards, community reports | Communication_Skill.md |
| 4. BENCHMARK | Cross-city comparison using Boston + San Francisco + Seattle + DC data | Comparing Boston to peer cities, learning from elsewhere | Benchmarking_Skill.md |
| 5. PERFORM | Results for America / PerformanceStat (CitiStat) | Budget × staffing × service outcomes: cost-per-outcome, workload-per-FTE, efficiency trends | Performance_Management_Skill.md |
Always read the relevant sub-skill file before beginning each phase.
Quick Decision Router
User Request
│
├─ "What data does Boston have on..." / "Help me define the problem"
│ → Read Problem_Framing_Skill.md → Run Phase 1
│
├─ "Analyze / run the numbers / what does the data show / is there an equity issue"
│ → Read Analytical_Skill.md → Run Phase 2
│
├─ "Write a memo / create a brief / make a dashboard / present these findings"
│ → Read Communication_Skill.md → Run Phase 3
│
├─ "Compare Boston to other cities / how does Boston rank / what works elsewhere"
│ → Read Benchmarking_Skill.md → Run Phase 4
│
├─ "Budget vs. performance / cost per outcome / workload per FTE / are we getting results / staffing efficiency / how much does it cost to / is the department understaffed / overtime analysis"
│ → Read Performance_Management_Skill.md → Run Phase 5
│
└─ "Full analysis / investigate / give me a recommendation / policy project"
→ Run all relevant phases in sequence
MCP Tool Reference — All Three Cities
Boston Open Data (Primary)
search_datasets(query) → Discover datasets by topic
get_dataset_info(dataset_id) → Find resource IDs and metadata
get_datastore_schema(resource_id)→ Get exact field names before querying
query_datastore(resource_id, filters={}, sort="", limit=100, date_range={})
San Francisco Open Data (Benchmarking — Socrata)
San Francisco Open Data:socrata__search_datasets(query)
San Francisco Open Data:socrata__get_dataset(dataset_id)
San Francisco Open Data:socrata__get_schema(resource_id)
San Francisco Open Data:socrata__query_dataset(resource_id, ...)
San Francisco Open Data:socrata__execute_sql(soql_query)
Seattle Open Data (Benchmarking — Socrata)
Seattle Open Data:socrata__search_datasets(query)
Seattle Open Data:socrata__get_dataset(dataset_id)
Seattle Open Data:socrata__get_schema(resource_id)
Seattle Open Data:socrata__query_dataset(resource_id, ...)
Seattle Open Data:socrata__execute_sql(soql_query)
DC Open Data (Benchmarking — ArcGIS)
DC Open Data:arcgis__search_datasets(query)
DC Open Data:arcgis__get_dataset(dataset_id)
DC Open Data:arcgis__query_data(dataset_id, ...)
DC Open Data:arcgis__get_aggregations(dataset_id, ...)
⚠️ ALWAYS confirm field names via schema before querying any dataset in any city.
Standard MCP Sequence (All Cities)
1. search_datasets("topic") → find dataset IDs
2. get_dataset_info("dataset-id") → find queryable resource IDs
3. get_datastore_schema(resource_id)→ confirm EXACT field names
4. query_datastore(resource_id, ...) → retrieve records
Boston 311 Schema Cheat Sheet (Critical)
The 311 system changed in October 2025. Field names differ:
| Concept | Legacy (2011–Oct 2025) | New System (Oct 2025+) |
|---|---|---|
| Open date | open_dt |
open_date |
| Close date | closed_dt |
close_date |
| Service type | type |
service_name |
| Department | department |
assigned_department |
| Neighborhood | neighborhood |
neighborhood (same) |
| On-time | on_time |
on_time (same) |
Key resource IDs: dff4d804-... (2024), 9d7c2214-... (Jan–Oct 2025), 254adca6-... (New System, Oct 2025+)
Cross-Phase Quality Standards
Rigor (J-PAL): Every claim is grounded in data or clearly labeled as interpretation. Confidence level stated. Limitations named, not buried.
Human-Centeredness (Bloomberg): Problem framed around people's lived experience. Recommendations are implementable by real city staff.
Inclusivity (GovLab): Equity lens applied. Plain-language versions exist. Feedback mechanisms included.
Transparency: Data sources cited with IDs. Methodology reproducible. Findings shareable as open knowledge.
Supporting Files in This Skill Set
| File | Purpose |
|---|---|
Problem_Framing_Skill.md |
Bloomberg methodology: scope, stakeholders, assumptions |
Analytical_Skill.md |
J-PAL methodology: descriptive → diagnostic → equity |
Communication_Skill.md |
GovLab/InnovateUS: memos, briefs, dashboards, engagement |
Benchmarking_Skill.md |
Cross-city comparison using San Francisco, Seattle, and DC data; includes Performance Management Benchmarking module |
TEMPLATES.md |
Fill-in-the-blank templates for 6 output types |
CHECKLISTS.md |
Pre-flight and review checklists for all phases |
PROMPTS.md |
Example prompts organized by phase and complexity |
REFERENCE.md |
Boston dataset directory, field names, cross-referencing |
Performance_Management_Skill.md |
Results for America / PerformanceStat: budget × staffing × outcomes efficiency analysis |
EXAMPLE-311-equity.md |
Complete worked example: 311 response equity analysis |
When Creating Documents
- Word docs (.docx): Also read
/mnt/skills/public/docx/SKILL.md - Presentations (.pptx): Also read
/mnt/skills/public/pptx/SKILL.md - Spreadsheets (.xlsx): Also read
/mnt/skills/public/xlsx/SKILL.md - Dashboards (React/HTML): Also read
/mnt/skills/public/frontend-design/SKILL.md
Source: sgarcese/Civic-Analytics-Agent-Workflow-Claude-Skill — distributed by TomeVault.