01
Search city data
Discover Syracuse datasets, inspect ArcGIS services, and query public records from both curated and raw sources.
Built for Syracuse & Onondaga County
Local data, finally ready to use.
Datacuse MCP gives AI assistants one reliable way to search Syracuse data, explore library services, and connect local questions to trusted public sources.
Open-source friendly · Local-first · No account required for public data
“Find a copy of The Color of Law and tell me which nearby OCPL branch has it.”
DATACUSE · search_ocpl_catalog
Searching catalog and branch availability…
Available at Petit Branch
1 copy available now. Catalog and branch records checked moments ago.
One connection, many systems
Ask local questions without hunting through portals.
Datacuse MCP keeps each source honest while making them feel like one coherent civic data layer.
02
Use library services
Find branches, events, catalog items, and availability—then continue into account-assisted workflows when needed.
03
Add trusted context
Connect local analysis to selected Census, labor, health, county, state, and federal sources.
Designed around real work
From plain-English question to a sourced answer.
The server handles the messy middle: finding the right system, understanding its shape, and returning a useful result.
1
Ask naturally
Use the language you would use with a colleague, not an API manual.
2
Route to the right source
Datacuse chooses a city, library, county, state, or federal tool.
3
Inspect and act
It resolves the relevant dataset, record, event, item, or workflow.
4
Confirm the result
You get a clear answer grounded in the original public system.
Zero-build local preview
Run this site in one command.
Clone or download the folder, then start the included dependency-free local server. The same files can be uploaded to any static host.
Terminal
node server.mjs
# Open http://localhost:4173Also works with npm start. No install, build, framework, or environment variables required.
Working examples
Built around useful local outcomes.
Read the launch essay: why Datacuse MCP exists
Public data is usually available in theory and painful in practice. Cities publish dashboards, ArcGIS services, CSV downloads, PDFs, portals, and one-off department pages. Libraries run separate catalogs, events systems, and reporting sites. The data is real, but using it across systems still takes too much manual work.
What makes it different
Datacuse MCP treats the local data ecosystem as connected but genuinely different systems. ArcGIS layers, catalog results, branch pages, events, annual statistics, and account actions are not forced into one fake abstraction. Shared discovery helps where it should; domain-specific tools handle the rest.
It is local and practical: find the right dataset, inspect records, locate a library item, and verify an outcome.
Why MCP fits the problem
Local public-data work is rarely a single API call. It is a chain: find the right source, understand it, resolve the right record, take the next action, and confirm the result. MCP gives an AI assistant a structured way to use the right capability at the right time.
What it is not
Datacuse MCP is not an official city or library product, a production civic data warehouse, or a substitute for secure multi-user authentication. It depends on upstream public systems that can change.
Where it could go
The same pattern can extend to more local sources, deeper state and federal integrations, and safer hosted workflows. It can work for any community whose public information is fragmented across portals, services, dashboards, and operational tools.
Public data becomes much more useful when it stops being a scavenger hunt.