Add a TlsMode read-only row and a None/StartTLS/SSL select to the SMTP
Configuration page edit form. New configs default to None; edits load
and persist the chosen mode through the repository.
Bundle G (#23 M7-T15): replace the temporary Admin-only gate on the Audit
Log surface with two new permission policies — OperationalAudit (read) and
AuditExport (bulk-export) — so the read path and the forensic-export path
can be delegated independently.
ScadaLink.Security
- AuthorizationPolicies: add OperationalAudit + AuditExport policy
constants; register them via RequireClaim with an explicit role allow-list
(OperationalAuditRoles, AuditExportRoles) so the role-to-permission
mapping is documented in one place.
- Default mapping: Admin and Audit roles grant both policies; AuditReadOnly
grants OperationalAudit only (read access without bulk export); Design
and Deployment grant neither.
ScadaLink.CentralUI
- AuditLogPage: switch the page-level [Authorize] to the OperationalAudit
policy and wrap the Export-CSV button in an AuthorizeView gated on
AuditExport so an OperationalAudit-only operator still sees the page +
filters but cannot trigger the CSV pull.
- ConfigurationAuditLog: switch from RequireAdmin to OperationalAudit so
both pages under the Audit nav group share the same gate.
- NavMenu: the Audit nav group now gates on OperationalAudit so the
section header + both child links match the per-page policies.
- AuditExportEndpoints: switch RequireAuthorization from RequireAdmin to
AuditExport — this is the authoritative gate; the AuthorizeView on the
button is just a UX affordance.
Tests
- New AuditLogPagePermissionTests covers the 5 brief-mandated cases plus
defence-in-depth for Admin-alone and AuditReadOnly users on the endpoint.
- SecurityTests: add policy-level coverage for the new role→permission
matrix (Theory rows pin every role/policy combination).
- AuditExportEndpointsTests: switch to AddScadaLinkAuthorization() so the
test host exercises the real production wiring under the new gate.
- AuditLogPageScaffoldTests: wrap the page render in a
CascadingAuthenticationState so the new in-page AuthorizeView resolves
the principal.
Adds three KPI tiles to the central Health dashboard for the Audit channel:
volume (rows in the last hour), error rate (Failed/Parked/Discarded over
total), and backlog (sum of SiteAuditBacklog.PendingCount across all sites).
Repo + service:
- IAuditLogRepository.GetKpiSnapshotAsync(window, nowUtc) — single aggregate
SELECT over the trailing window returning total + error counts; nowUtc is
optional for production callers and pinned by integration tests against the
shared MSSQL fixture so the global counts are deterministic.
- AuditLogQueryService.GetKpiSnapshotAsync() — composes the repo aggregate
with a sum of SiteAuditBacklog.PendingCount read from ICentralHealthAggregator.
- AuditLogKpiSnapshot record in Commons/Types/.
UI:
- New AuditKpiTiles Blazor component (Components/Health/) — three Bootstrap
card-tiles, click navigates to /audit/log with the matching pre-filter.
- Health.razor wires the tiles in alongside the existing Notification Outbox
KPIs; LoadAuditKpis() runs on every 10s refresh tick and degrades to em
dashes + inline error if the query fails.
- AuditLogPage extended to parse ?status= so the error-rate tile drill-in
(?status=Failed) auto-loads the grid.
Tests:
- AuditLogRepositoryTests: GetKpiSnapshotAsync mixed-status + empty-window
cases against the MSSQL migration fixture.
- AuditLogQueryServiceTests: forwarding + backlog composition; sites with
null SiteAuditBacklog contribute zero.
- AuditKpiTilesTests: 9 bUnit tests covering tile render, error-rate maths
with safe zero-events handling, em-dash unavailable path, click-through
navigation, and warning/danger border thresholds.
- HealthPageTests: new Renders_AuditKpiTiles_WithValues plus IAuditLogQueryService
stub registration in the constructor so existing outbox tests still pass.
- AuditLogPageScaffoldTests: ?status=Failed auto-load + unknown status drop.
Adds "Recent audit activity" deep links from four edit/detail pages into
the central Audit Log, each with a pre-filter encoded in the query string
that the Audit Log page (Bundle D0) now parses on initialization:
- External Systems (Design/ExternalSystemForm) → ?target={Name}
- API Keys (Admin/ApiKeyForm) → ?actor={Name}&channel=ApiInbound
- Sites (Admin/SiteForm) → ?site={SiteIdentifier}
- Instances (Deployment/InstanceConfigure) → ?instance={UniqueName}
The link is suppressed on create/new flows where there is nothing to
drill into yet. Instance is UI-only on the filter bar (the repository
filter contract has no instance column), so the page-side prefill threads
through the InitialInstanceSearch seam on AuditFilterBar.
Site Calls (#22 M7-T11) drill-in is DEFERRED: the Central UI does not
yet host a Site Calls listing page, per M3 reality notes. Add the
drill-in when that page lands.
#23 M7-T12
Implements Bundle C (M7-T4 through M7-T8) of the Audit Log #23 M7
Central UI work: a right-side off-canvas drawer that opens from
AuditResultsGrid row clicks and renders one AuditEvent in full.
Cohesive single-component delivery:
- Read-only fields stacked (form-layout memory): Channel/Kind, Status,
HttpStatus, Target, Actor, Source* provenance, CorrelationId,
OccurredAtUtc, IngestedAtUtc, DurationMs.
- Channel-aware body renderer: DbOutbound {sql, parameters} payloads
render a code-block with CSS-only .language-sql class plus a
parameter <dl>; other channels JSON-pretty-print when parseable and
fall back to verbatim <pre>.
- Redaction badges on Request/Response when the body contains the
<redacted> or <redacted: redactor error> sentinels.
- Copy-as-cURL (API channels only) builds a curl command from Target
+ optional {method, headers, body} RequestSummary JSON and writes
it via navigator.clipboard.writeText.
- Show-all-events drill-back navigates to /audit/log?correlationId={id}
when the event carries a CorrelationId.
- Close button + backdrop-click both raise OnClose.
AuditLogPage wires Event/IsOpen/OnClose; row clicks now open the
drawer (HandleRowSelected pins _selectedEvent + _drawerOpen=true).
11 bUnit tests cover field rendering, JSON pretty-print, verbatim
fallback, SQL block, conditional buttons, redaction badges,
navigation drill-back, and clipboard interop. No third-party UI
libraries: Bootstrap offcanvas + scoped razor.css only.
Adds the results grid + query facade for the central Audit Log page
(#23 M7-T3):
* IAuditLogQueryService / AuditLogQueryService — CentralUI facade over
IAuditLogRepository.QueryAsync so the grid can be tested with a stubbed
query source. Default page size is 100; callers can override per call.
* AuditResultsGrid.razor + .razor.cs — Blazor Server component (Bootstrap
only, no third-party UI libs). Renders the 10 columns from
Component-AuditLog.md §10 (OccurredAtUtc, Site, Channel, Kind, Status,
Target, Actor, DurationMs, HttpStatus, ErrorMessage). Keyset-paged via
the last visible row's (OccurredAtUtc, EventId) as the cursor; Next-page
button disabled when the current page is short (no count query). Row
clicks emit OnRowSelected(AuditEvent) for Bundle C's drilldown drawer.
Status badges are colour-coded (Delivered=green; Failed/Parked/Discarded
=red; other=gray). Error messages truncated to 80 chars with full text
on hover.
* Column model framework: a ColumnOrder [Parameter] reorders columns by
stable string keys; unknown keys are dropped. M7 scope decision (in the
class doc): the framework is in place but drag-reorder / resize UX is
not implemented — M7.x can add persisted-per-user reordering without
rewriting the column model.
* AuditLogPage wired: hosts AuditFilterBar + AuditResultsGrid, threads
the filter through and stubs OnRowSelected for Bundle C.
* AuditLogQueryService registered as scoped in AddCentralUI.
* Tests: 4 grid bUnit tests (10 columns rendered, next-page cursor
carries last row, row click raises callback, badge classes for
Failed vs Delivered), 2 service tests (filter+paging pass-through,
default page size of 100). AuditLogPageScaffoldTests updated to
provide the new ISiteRepository + IAuditLogQueryService stubs the
page now resolves.
Adds the filter bar for the central Audit Log page (#23 M7-T2):
* AuditQueryModel — UI binding model with chip-style multi-select state for
Channel/Kind/Status/Site, a Channel→Kind narrowing map (CachedSubmit and
CachedResolve appear under both ApiOutbound and DbOutbound per
Component-AuditLog.md §4), time-range presets (5min/1h/24h/Custom),
free-text Instance/Script/Target/Actor searches and an Errors-only toggle.
Collapses to the single-value AuditLogQueryFilter on ToFilter(utcNow);
multi-select chips take the first selected per dimension and the
Errors-only toggle pins Failed when Status chips are empty (chip-set wins
otherwise) — documented Bundle B scope decision.
* AuditFilterBar.razor + .razor.cs — Blazor Server component (Bootstrap
only, no third-party UI libs). Renders the 10 spec elements plus the
Errors-only toggle, populates Site chips from ISiteRepository at
initialisation, exposes [Parameter] EventCallback<AuditLogQueryFilter>
OnFilterChanged and an optional NowUtcProvider seam for time-window tests.
* AuditFilterBarTests — 5 bUnit tests pinning element presence, Apply
callback payload, Channel→Kind narrowing, Errors-only toggle precedence
and the LastHour time-window collapse.
Adds the central-side Audit Log page scaffold at /audit/log (M7-T1) and
introduces a new Audit nav group (M7-T9) that hosts both the new page and
the renamed Configuration Audit Log. The page body is intentionally a
heading + two placeholders — Bundle B will land the AuditFilterBar and
AuditResultsGrid behind them.
The Audit nav group sits between Monitoring and the per-user footer; both
items inside are Admin-only, so the section header lives inside the
RequireAdmin AuthorizeView (non-admins see no orphan header).
bUnit scaffold tests pin the page heading, the section header order, and
the two child links; the existing 338 CentralUI tests continue to pass.
The script-analysis sandbox Notify surface was stale after the Notification
Outbox change: SandboxNotifyTarget.Send returned Task<NotificationResult> and
there was no Status method, while production NotifyTarget.Send returns
Task<string> (a NotificationId) plus NotifyHelper.Status. A script that
test-ran cleanly in the sandbox would not compile against the real site
runtime.
- Move the NotificationDeliveryStatus record from ScadaLink.SiteRuntime.Scripts
into ScadaLink.Commons.Messages.Notification so both production and the
CentralUI sandbox reference the exact same type (CentralUI does not, and
should not, reference SiteRuntime). Production NotifyHelper.Status is
otherwise untouched.
- Rewrite SandboxNotifyHelper/SandboxNotifyTarget to be a signature-faithful
no-op fake: Send returns Task<string> (a fake NotificationId), Status returns
Task<NotificationDeliveryStatus>. Production now enqueues into the site S&F
engine, which has no central-side equivalent in the sandbox, so the fake no
longer carries an INotificationDeliveryService.
- Add script-analysis tests proving a script using the new Notify shape both
diagnoses clean and runs in the sandbox.
The template script editor had no input for MinTimeBetweenRuns, so a
WhileTrue trigger configured through the UI always saved a null interval
and degraded to a single edge fire. The Add/Edit Script modal now has a
"Min time between runs" number+unit (ms/sec/min) field.
- Visible only for ValueChange / Conditional / Expression triggers — the
auto-firing triggers MinTimeBetweenRuns throttles. Hidden for Interval
(its own period is the cadence), Call (invoked explicitly, never
throttled), and None.
- For a WhileTrue Conditional/Expression trigger the field is labelled as
the re-fire interval and shows a warning while it is blank.
- Wired through the new-script and edit-script save paths (edit previously
only preserved the existing value, never let the user change it).
New DurationInput helper does the TimeSpan <-> number+unit conversion;
ScriptTriggerConfigCodec.SupportsMinTimeBetweenRuns classifies trigger
types. Both TDD'd — 21 new tests. CentralUI suite 316 green; verified
end-to-end in the browser (visibility per trigger type, WhileTrue warning,
save/reload round-trip).
Conditional and Expression script triggers gain an optional `mode` field
in their TriggerConfiguration JSON:
- OnTrue (default): unchanged edge/per-change firing. An absent mode field
parses as OnTrue, so every existing trigger config behaves identically.
- WhileTrue: fires on the false->true edge, then re-fires on a periodic
timer while the condition holds; stops on the true->false edge. The
re-fire cadence is the script's MinTimeBetweenRuns; with none configured
the trigger degrades to a single edge fire and logs a warning.
ScriptActor tracks condition truth state and manages a dedicated
"whiletrue-trigger" timer. ScriptTriggerConfigCodec and ScriptTriggerEditor
round-trip the mode and expose an OnTrue/WhileTrue selector for the two
trigger kinds. Design: docs/plans/2026-05-18-whiletrue-trigger-mode-design.md
Tests: 7 ScriptActor runtime tests (edge fire, timer re-fire, stop,
re-arm, no-MinTimeBetweenRuns degrade, OnTrue regressions) + 14 codec /
editor tests. SiteRuntime suite 206 green, CentralUI suite 295 green.
DiffDialogTests.SetupBodyLockInterop registered bUnit SetupVoid planned
invocations that were never completed; DisposeAsync_WhileOpen awaited
DiffDialog.DisposeAsync -> TryUnlockBodyAsync -> InvokeVoidAsync on one of
them, suspending the test forever so the test host never exited (regression
from the CentralUI-023 catch-narrowing). SetupBodyLockInterop now uses Loose
JSInterop mode. Also dispose the leaked WebApplication instances in the Auth
tests (FileSystemWatcher + ConsoleLoggerProcessor threads) and the extra
ServiceProvider in the DebugView tests. Suite now runs 281 tests in ~7s and
exits cleanly.
A heartbeat-registered site that has never sent a full report now has
LastReportReceivedAt = null instead of the year-0001 sentinel. TimestampDisplay
accepts DateTimeOffset? and renders null as a placeholder ('awaiting first
report') rather than a ~2000-year-stale date. Cross-module: HealthMonitoring +
CentralUI.
ScriptAnalysisService.RunInSandboxAsync compiled and executed arbitrary
user C# in the central host process with no trust-model enforcement — the
forbidden-API set was only a Monaco editor diagnostic. A Design-role user
could run System.IO/Process/Reflection/network code on the central node.
Added a Roslyn semantic gate (EnforceTrustModel) invoked after compilation
and before script.RunAsync, and on nested shared scripts in callSharedFunc;
a script referencing any forbidden API is rejected before it runs.
Reworked FindForbiddenApiUsages: it now resolves every identifier against
the semantic model and checks types and members, so a fully-qualified call
(System.IO.File.WriteAllText) is caught — the pre-fix check only inspected
the leftmost identifier and missed that shape. This is a static semantic
gate, not a process sandbox.
Adds gate regression tests that fail against the pre-fix code, plus a
clean-script test guarding against over-blocking.
Move all package versions into Directory.Packages.props so every project
resolves a single consistent version. Consolidates the Roslyn packages
(Microsoft.CodeAnalysis.CSharp.Scripting/Workspaces) onto 5.0.0, which
resolves the pre-existing NU1608 version-skew error in the test projects.
Six tests asserted DoesNotContain(SCADA004/SCADA005) or an empty InlayHints
result — all pass for the wrong reason now that those diagnostics and the
positional InlayHints were removed in the analyzer realignment. They also
used the obsolete top-level CallScript syntax. Removed.
Two stale references blocked compilation: the DataConnection page tests
still pointed at Components.Pages.Admin (the pages moved to .Design), and
ScriptAnalysisServiceTests constructed ScriptAnalysisService without the
IServiceProvider parameter. The project now compiles.
Adds a new HiLo alarm trigger type with four configurable setpoints
(LoLo / Lo / Hi / HiHi). Each setpoint carries an optional priority,
deadband (for hysteresis), and operator message. The site runtime emits
AlarmStateChanged with an AlarmLevel field so consumers can differentiate
warning vs critical bands.
Plumbing:
- new AlarmLevel enum + AlarmStateChanged.Level/Message init properties
- AlarmTriggerEditor (Blazor) gets a HiLo render with severity tinting
- AlarmTriggerConfigCodec extracted from the editor for testability
- sitestream.proto carries level + message over gRPC
- SemanticValidator enforces numeric attribute, setpoint ordering,
non-negative deadband
- on-trigger scripts get an Alarm global (Name/Level/Priority/Message)
so notification routing can branch by severity
- per-instance InstanceAlarmOverride entity + EF migration + flattening
step + CLI commands; HiLo overrides merge setpoint-by-setpoint, binary
types whole-replace
- DebugView shows a Level badge + per-band message tooltip
- App.razor auto-reloads on permanent Blazor circuit failure
- docker/regen-proto.sh automates the proto regen workflow (the linux/arm64
protoc segfault means generated files are checked in for now)
Replace raw-JSON text inputs with rich UI: script parameter/return types use
a JSON Schema builder (SchemaBuilder + JsonSchemaShapeParser, with a migration
to convert existing definitions); alarm trigger config uses a type-aware
editor with a flattened attribute picker (AlarmTriggerEditor). AlarmActor
gains optional direction (rising/falling/either) on RateOfChange triggers.
Phases 3+4 of the script-scope rollout. Wires the runtime accessors
landed in efba01d through to Monaco completion, diagnostics, and
hover.
New analyzer surface in ScriptAnalysisService:
String-literal completion contexts (added to TryStringLiteralCompletions):
Attributes["..."] -> SelfAttributes
Children["..."] -> composition names
Children["X"].Attributes["..."] -> child template's attributes
Children["X"].CallScript("...") -> child template's scripts
Parent.Attributes["..."] -> parent template's attributes
Parent.CallScript("...") -> parent template's scripts
Diagnostics:
SCADA006 Attribute "Typo" is not declared on {this template,
child composition 'X', the parent}. (warning)
SCADA007 Composition "Unknown" is not declared on this template.
(warning)
CallShared / CallScript snippet-expansion now routes through the
child / parent shape catalogs when invoked on Children["X"] /
Parent — picking a child script accepts `Sample", ${1:count})`.
Contract additions:
- AttributeShape (Name, Type) record
- CompositionContext (Name, Attributes, Scripts) record
- SelfAttributes / Children / Parent fields on DiagnoseRequest,
CompletionsRequest, HoverRequest, SignatureHelpRequest
ScriptHost (analyzer-side globals) gains stub AttributeBag /
ChildrenBag / CompositionBag types so Roslyn doesn't emit CS0103 on
Attributes / Children / Parent. The stubs are never invoked — only
their signatures are read by the analyzer's compilation pass.
MonacoEditor.razor exposes SelfAttributes / Children / Parent
parameters; GetContext returns them; monaco-init.js forwards all
three on completion / hover / signature-help / diagnostics requests.
TemplateEdit fetches each composition's resolved child template
shape via GetTemplateWithChildrenAsync, and queries GetAllTemplatesAsync
for any single parent that composes the open template. Multi-parent
or no-parent → Parent is suppressed.
11 new xUnit tests on the new completion / diagnostic paths. Total:
149 -> 159.
Browser-verified via curl:
- Children["..."] suggests composition names
- Attributes["..."] suggests attributes with type detail
- Attributes["Typo"] squiggles SCADA006
- Children["Unknown"] squiggles SCADA007
- No spurious CS0103 on the new accessors
Hover, signature help, and inlay hints for the new accessors keep
working because they reuse the same dispatch logic.
