xRMProducts.HealthScan 0.6.0.1

xRM Health Scan — xRM PRODUCTS

Local release candidate 0.6.0.1. NuGet ID xRMProducts.HealthScan, plugin assembly xRMProducts.HealthScan.dll, base namespace xRMProducts.HealthScan. Internal evaluation only; nothing published. Final terms explicitly permitting free use are still pending. No open-source license selected.

Six real data-quality checks: missing Lead country, E.164 phone format, possible Account duplicates, Contacts without email/phone, postal/country formats, and open opportunities without an estimated close date. Includes evidence, transparent check/category scores, HTML/PDF reports, cancellation and connection isolation. Five bounded security checks review role privileges and audit configuration. Field security is a neutral configuration inventory (secured columns, profiles, permissions and visible assignments), without inferred risk findings or scores; missing sources produce explicit partial/unassessed states, with no invented scores. Reads use the existing XrmToolBox connection only. No product solution, paid product license or write access is required.

Requires XrmToolBox 1.2025.10.74+, .NET Framework 4.8 and matching Microsoft Edge WebView2 Evergreen Runtime. Copy the package's lib/net48/Plugins contents into an isolated test profile's Plugins directory, preserving the asset subfolder. Do not copy SDK/host DLLs from a build output directory. Restart XrmToolBox.

After cancellation: Scan abgebrochen. After connection change: Verbindung geändert – bitte erneut scannen. Reports stay disabled until a new successful scan; late callbacks from the old connection are discarded.

Documentation: installation and live acceptance, checks and scoring, test results, license audit. Host setup documentation: https://www.xrmtoolbox.com/documentation/for-users/ Product owner: https://www.xrmproducts.com/ Dedicated public Health Scan documentation remains pending.

Release notes – 0.6.0.1

Model 2: five independent category scores, linear relative user/root-role populations and hierarchical audit configuration. Security is their equal mean; Overall keeps DQ/security at 50% each. DQ and findings unchanged. Active non-admin application/user denominators explicitly captured; unknown/empty/partial bases separately explained. Model 2 is a versioned product heuristic, not a Microsoft standard or security certification, and is not directly comparable to Model 1. DE/EN category formulas in UI/HTML/PDF. Read-only, stale-result guards and export selection retained. Frozen DEV/SENEC comparison uses known aggregates only; historical non-admin app denominator missing, no new scan. No new dependency, no publication.

Previous release behavior (0.5.0.1)

Stable compact DE/EN desktop navigation. Shared summary/detail selector in preview and export dialog; HTML/PDF from either tab, distinct type/language filenames. Implemented capped security deductions and equal-weight overall score with explicit coverage and provisional partial scoring. User checks count only systemuser.isdisabled = false; disabled users are neutral exclusions, no findings, examples or deductions. Role configuration remains independent. Read-only and stale-result safeguards retained. Model and examples: SCORING.

Previous functional behavior (0.2.2.1)

Binding product/package/assembly identity. Five data-quality checks with real results and bounded evidence, score/coverage explanations, precise cancellation and connection statuses, export locks, factual product hints and third-party license notices. Actual XrmToolBox/Dataverse security acceptance and final distribution-license review remain pending. Do not publish.

0.2.1.1 added the supplied original X logo, persistent responsive product header, per-table execution coverage, and a German A4 executive report. Real findings come first; missing scores state their reason; Lead country completeness is a field-specific metric. Twelve synthetic viewport/zoom cases and two one-page PDF regression cases passed; these are not live Dataverse or Windows DPI acceptance.

Language and scrolling

English is the first-run default. Select English, Deutsch or Automatic (host). Automatic uses the host UI thread culture captured at startup, with English fallback; Dataverse base language does not select the UI language. The local preference applies to UI, results, status, HTML and PDF. Switching language preserves scan results and completion time without querying Dataverse again.

Fixed header and stationary navigation; independent middle/detail/report scrolling. All navigation categories remain reachable through a compact selector at small sizes. Check counts and actual finding counts are distinct. Records read are shown separately from value denominators for each check. New host acceptance follows local file import.

Previous release behavior (0.4.0.1)

Adds the approved open-opportunity rule: only open records enter the denominator; closed records are separately excluded; no open records means no score. Adds a neutral field-security configuration inventory for selected tables, including secured columns, scoped profiles, permission flags, optional unmasked-read flags and visible user/team assignments. Neither profiles nor assignments create risk findings or scores. Missing sources or visibility remain explicit. DE/EN UI, paginated configuration details, HTML/PDF reports and cancellation/connection isolation use the existing read-only connection. Six DQ checks, five security checks and one configuration inventory are implemented. The previous precise scores, scoped evidence, name-only public views and separate executive/detail reports remain. No new third-party dependency. Local test only; nothing published. See docs/TEST-RESULTS.md for actual tests and open live/release acceptance.

Final Version 1 scope

Live progress shows localized phases, elapsed time, actual query rows read and fully completed checks. Rows can be read by several checks and are not a count of unique environment records. The progress bar is indeterminate: no fabricated percentages or remaining time. Total scan duration survives language changes and is included in reports. Cancel remains reachable in narrow windows; old run progress cannot overwrite a new run.

Supported account/contact/lead/opportunity evidence offers Open in Dynamics. The URL uses the captured connection environment, logical table and known record ID and is launched in the external system browser. No SDK read/write is needed for opening. Unknown environment URLs, unsupported record views and unknown or stale evidence do not get links. No personal record links are exported as samples. No scan comparison feature; use the existing exports. Nothing published.