ClearParcel GIS Data Watchtower

Explore the availability, freshness, and field population of Minnesota public GIS data.

Explore the live public dashboard · User guide · Documentation index

The current production release is v0.3.12, deployed November 3, 2026 to the public and private dashboards and the Cloud Run jobs, including the weekly digest. See STATUS.md and the release record. Starting with v0.3.10, each deployed version is tagged and published as a GitHub Release; v0.3.0 through v0.3.9 are recorded but not tagged (see the release policy). Alert delivery to the operator email channel is enabled; the five-cycle observation is in progress.

Operators can use the offline triage runbook to review saved provider-run evidence without triggering new checks or notifications. The alert-routing rehearsal prepares private failure, overdue-check and recovery message previews, plus a durable delivery ledger whose transport stays disabled by default. The authority reconciliation runbook checks supplied monitoring-approval evidence without activating sources. The isolated restore rehearsal exercises synthetic aggregate recovery and concurrent publishing in temporary local storage.

Actual public dashboard, captured October 12, 2026. Counts, check times, and field-population percentages change as new observations arrive.

What is Watchtower?

ClearParcel GIS Data Watchtower is an open-source Python monitoring and reporting application. Its public website helps county staff, GIS professionals, researchers, and residents understand where Minnesota public geospatial data is available and what the latest recorded checks show.

It monitors selected public ArcGIS, WMS/WFS, and other supported GIS services, records changes and data-quality signals, and publishes a sanitized, read-only dashboard. It is not a property-ownership search, a complete inventory of every GIS service, an official source-data repository, or a certification of a county's compliance with a data standard.

What you can explore

Public view	What it helps answer
Overview	How many service entries are monitored? Which had warnings? How many Minnesota counties have parcel observations?
Minnesota Counties	What access, statewide availability, county contacts, and monitoring evidence are recorded for a county?
MN GAC	How frequently are standard parcel, address-point, and road-centerline fields populated in available statewide data?
Data Sources	Which public GIS sources are checked, and what were their latest recorded statuses and observation times?
Exports	How can I reuse published summary evidence as JSON, CSV, or Excel without downloading private operational data or raw parcel-owner records?

Four distinctions matter:

• Field population is not a compliance grade. A blank Conditional, If Available, or Optional field can be valid. Mandatory-field and all-field statistics answer different questions.
• No data does not mean zero. A county absent from a particular statewide open layer is not automatically missing data at the county level or failing a standard.
• County scan labels use the saved observation. Shared source-schema metadata accompanies county-specific measurements; a measured zero remains scanned, while an unscanned or unavailable result is not assigned a population percentage. See county field interpretation.
• Access, monitoring, and publication are different. A county may publish data directly, appear in MnGeo's opt-in statewide compilation, be observed through either or both paths, or await access-policy clarification.

See the illustrated public user guide for step-by-step instructions and examples.

Explore a county

1. Open the public dashboard.
2. On Overview, select a county on the map or with Find a county. The adjacent panel provides recorded monitoring and field-population details.
3. Select Open county profile for source, contact, and access evidence, or Explore county MN GAC fields for field-by-field results.
4. Use MN GAC to switch between Parcels, Address Points, and Road Centerlines and compare counties.
5. Use Export or page-specific export controls to obtain a dated, reusable summary.

Illustrative county selection as of October 12, 2026. Direct-feed observations are not county-wide compliance findings.

Data, freshness, and responsible use

• Watchtower separates source health, provider check freshness, and public snapshot publication freshness. A recently published page does not necessarily mean an external provider was just checked.
• The approved cloud provider checks run daily at 5:00 a.m. America/Chicago. A separate sanitized public publisher runs every 10 minutes. Always consult observation timestamps.
• The public site exposes derived counts, field population, statuses, and attribution—not parcel-owner records, private credentials, operational provider endpoints, or diagnostic fingerprints.
• The system uses bounded, respectful requests. It does not scrape human-facing property-search websites or bypass provider access restrictions. Public reachability does not grant redistribution rights.

Read the MN GAC methodology, county access methodology, provider-use guidance, and public/private design.

Install and run your own monitor

Requires Python 3.12 or later:

python -m venv .venv
# Activate .venv for your platform, then:
python -m pip install -e .

For optional Google Cloud Storage support:

python -m pip install -e ".[gcs]"

The included config/example_sources.json is intentionally provider-neutral. Configure only services you are authorized to monitor. See the configuration reference for sources, storage, request limits, authentication, and environment variables.

watchtower --config config/example_sources.json check
watchtower --config config/example_sources.json status
watchtower --config config/example_sources.json dashboard --host 127.0.0.1 --port 8765

A local dashboard defaults to private/loopback access; Internet-facing deployments require appropriate authentication and TLS controls. The hosted public view is a separate read-only presentation of sanitized data, not the private operational dashboard.

Developers can use the documentation index, contribution guidelines, and cloud/offline development guide. CI covers Windows and Ubuntu, Python 3.12–3.14, container smoke tests, dependency auditing, CodeQL, and image scanning.

Releases and project status

The current production release is v0.3.7 on the public and private dashboards (deployed October 23, 2026); the provider and public-publisher jobs remain on v0.3.6 (deployed October 20). Each component's current release record lists its images, revisions and rollback targets. The footer displays only the version number; the About section explains the open-source ClearParcel mission, disclaimer, and GitHub link.

• v0.3.7 dashboard release — current dashboard images, revisions, and rollback to v0.3.6 revisions
• v0.3.6 coordinated release — current provider and publisher job images
• Validated operational status — current monitoring and remaining gates
• GitHub releases — versioned summaries; the latest published there is v0.2.2, so the release records above are authoritative for v0.3.x
• Documentation index — user guides, methodology, engineering references and dated history
• Original v0.2.0 release and October 11 correction — retained history

Earlier merged PR #69 static-output/provider retry corrections also ship in the v0.2.2 application image. The coordinated v0.2.2 release updated both dashboard services and the existing provider/publisher job images without changing source registries, security settings, or schedules.

Historical: the v0.3.0 rollout (October 14) updated both dashboard services and the provider and public-publisher job images. Monitoring schedules, source registries and security boundaries were preserved. The v0.2.3 rollout had updated only the dashboard services, so v0.2.2 was then the job rollback baseline.

This is a pre-1.0 project. Open issues may represent documented provider-policy, privacy, and observation gates rather than defects.

Contribute, ask questions, and report problems

Contributions and corrections are welcome. Use GitHub Issues for reproducible bugs, public-data corrections, and enhancements, and follow CONTRIBUTING.md.

Do not report vulnerabilities in public issues. Follow the private reporting process in SECURITY.md.

ClearParcel GIS Data Watchtower is licensed under the MIT License. Source agencies remain authoritative for their own datasets, definitions, release policies, and attribution requirements.