Skip to main content
Glama
Sign In
seattle-open-data

seattle-open-data

by briandconnelly
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Remote

seattle-open-data

CI Python License: MIT

An agent-friendly MCP server for Seattle Open Data (Socrata SODA2).

WARNING

This project is deprecated and no longer maintained. It has been superseded by opendata-atlas, which covers Seattle Open Data and other Socrata portals. Please switch to it. 0.2.1 is the final release.

Status: deprecated (0.2.1). All four tools work against the live portal, with curated guides for seven flagship datasets. See evals/results for the agent evaluation.

Tools

Tool

Purpose

seattle_data_search_datasets

Find datasets by topic

seattle_data_describe_dataset

Columns, row count, time coverage, place fields, geo columns

seattle_data_query_dataset

Bounded reads and aggregates (SoQL clauses or raw SoQL)

seattle_data_get_reference

Capability summary, SoQL help, guides, error catalog

Related MCP server: data-seattle

Install

The server runs over stdio. The uvx commands below need uv.

Claude Desktop: download seattle-open-data-<version>.mcpb from the latest release and open it with Claude Desktop.

Claude Code:

claude mcp add seattle-open-data -- uvx seattle-open-data

Codex:

codex mcp add seattle-open-data -- uvx seattle-open-data

Any other MCP client: run uvx seattle-open-data as a stdio server:

{
  "mcpServers": {
    "seattle-open-data": { "command": "uvx", "args": ["seattle-open-data"] }
  }
}

Optional app token: set SOCRATA_APP_TOKEN to use a Socrata application token, for example claude mcp add seattle-open-data -e SOCRATA_APP_TOKEN=... -- uvx seattle-open-data. The bundle asks for it in Claude Desktop's settings. A rejected token is reported as app_token_rejected, never silently dropped.

Develop

From a checkout:

uv sync
uv run seattle-open-data                                  # run the server over stdio
uv run pytest
uv run ruff check && uv run ruff format --check && uv run ty check
UPDATE_SNAPSHOTS=1 uv run pytest tests/test_contract.py   # after an intended tool-surface change
uv run pytest --run-integration -m integration            # live tests against data.seattle.gov
uv run python evals/run.py [--host claude|codex] <out_dir> [task_id ...]
uv run python evals/report.py <out_dir> --answers

To build the MCPB bundle locally (needs the mcpb CLI): make mcpb.

Release

Tag vX.Y.Z after bumping the version: the release workflow builds and smoke-tests the wheel, sdist, and .mcpb, publishes to PyPI after maintainer approval, and attaches all three to a GitHub Release. See docs/releasing.md for the steps, the TestPyPI rehearsal, and one-time setup.

License

MIT

Available Tools

4 tools
seattle_data_describe_datasetSeattle Data Describe DatasetA
Read-only

Decide whether and how to query a dataset: columns, live row count, data_coverage (actual time range), place fields for "in " questions (with reliability), and geo columns for near.

Call before querying a dataset whose columns you do not know. For area questions, use detail="full". Filter a place field on its value, never its label; an area may span several values. Example: {"dataset_id": "kzjm-xkqj", "detail": "full"}. Typically 1-3 s; fails with query_timeout after 30 s.

Inputs: dataset_id (required), detail, sample_rows, column_query. Returns: title, source, columns, guide_uri, +11 more

seattle_data_get_referenceSeattle Data Get ReferenceA
Read-only

Read the server's reference: capability summary, SoQL help, dataset guides, and the full error catalog. Call with no arguments for the index.

Example: {"topic": "soql", "item_id": "dates"}.

Inputs: topic, item_id. Returns: uri, body, title, topic, +4 more

seattle_data_query_datasetSeattle Data Query DatasetA
Read-only

Run a bounded read or aggregate on one dataset; paging stops at 10,000 rows.

For counts and breakdowns, use count() with group_by, not row paging. With cursor, send only dataset_id, cursor, and optionally page_size or detail. For all rows, give the user download_url (detail="full") instead of paging. Example: {"dataset_id": "kzjm-xkqj", "select": ["type", "count() AS n"], "group_by": ["type"], "order_by": ["n DESC"]}. Typically 1-5 s; fails with query_timeout after 30 s.

Inputs: dataset_id (required), mode, near, soql, +9 more. Returns: rows, source, columns, has_more, +13 more

seattle_data_search_datasetsSeattle Data Search DatasetsA
Read-only

Find Seattle datasets by topic. Call first when you lack a dataset id.

Results are candidates: match_evidence lists the query words each one's metadata lacks. Only queryability "supported" can be described or queried; a tabular search also returns up to 3 external_matches, which cannot. Example: {"query": "fire 911"}. Typically 1-3 s.

Inputs: query, cursor, detail, category, +2 more. Returns: source, datasets, has_more, suggestion, +4 more

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Deploy Server
    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    TDQS

    A4.6/5.0

    Scored across 4 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose: search for datasets, describe a specific dataset's metadata, access server-wide reference documentation, and query a dataset. The descriptions clearly delineate when to use each, with no overlapping functionality.

    Naming Consistency5/5

    All tools follow the consistent pattern seattle_data_<verb>_<noun> (search_datasets, describe_dataset, get_reference, query_dataset). The verbs are distinct but appropriate, and the naming is predictable and readable.

    Tool Count5/5

    Four tools are well-scoped for a read-only open data query service, covering discovery, metadata inspection, reference, and query execution without redundancy. Each tool earns its place.

    Completeness4/5

    The toolset covers the core workflow: search datasets, describe schema, query data, and access reference help. A minor gap is the lack of a tool to list all datasets without a topic query, but this is a minor limitation in an otherwise complete surface for the domain.

    Maintenance

    ActivityMaintained
    ResponsivenessOutdated data

    Resources

    Related MCP Connectors

    Related MCP Servers

    • DataSF MCP

      pipeworx-io
      A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI agents to query and retrieve San Francisco open data from data.sfgov.org via the Socrata SODA API.
      164 npm
      MIT

    • mcp-data-cincinnati

      pipeworx-io
      A
      license
      Not graded
      quality
      B
      maintenance
      Provides access to Cincinnati open data via the Socrata SODA API, allowing users to query datasets using natural language or direct tool calls.
      153 npm
      MIT

    • mcp-data-orlando

      pipeworx-io
      A
      license
      Not graded
      quality
      B
      maintenance
      Enables searching and querying Orlando Open Data datasets via Socrata SoQL, providing access to metadata and data rows through natural language or direct tool calls.
      147 npm
      MIT