Security finding (CWE-367): the activation wrote the approved row first and the OnboardedBy
admin row a few storage round-trips later. In between the tenant was activated with no
members, and auth/me seated any roleless caller of that moment as Admin; every member of the
customer's Entra tenant could poll auth/me across the auto-approve moment.
- TenantApprovalService: uncached pre-check, then the OnboardedBy Admin row (conditional
insert, best-effort, fresh config read), then the activation row (still the arbiter), then
the welcome mail. The two tables cannot share a batch, so "before" is the guarantee.
- AuthFunction.BuildAuthResult: with a promotable OnboardedBy the first-user auto-admin goes
to that UPN only; other roleless callers stay roleless. Without one (rows from before the
field existed) the first sign-in still becomes the admin. The grant logs at Warning.
- Both first-admin writers use IAdminRepository.TryAddTenantMemberAsync (AddEntityAsync,
409 = left untouched) instead of an upsert; TenantAdminsService.TryAddTenantAdminAsync
replaces AddTenantAdminAsync. PreviewWhitelistService.IsApprovedFreshAsync is the pre-check.
- Dead after this: IsTenantAdminAsync on the service, the repository interface and the table
repository, plus its test.
- Tests pin the write order, the designated-member guard (unit and auth/me end to end) and the
conditional write; each part was mutated back to the old code and failed its tests.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
After auto-activation, a third of new tenants' first admins were answered by
auth/me without a role and sent from the activation page to /progress. The
auto-approve worker promotes the onboarding member on another instance, so the
portal's caches never hear of it: auth/me combined a cached "no row" (refreshed
1.5 s after every activation poll by feedback/status) with a fresh member list,
and neither the role nor the first-user auto-admin applied.
- TenantAdminsService caches only an enabled membership; a missing or disabled
row is read on every call. IsTenantAdminAsync reads through the same cache.
- auth/me takes the caller's row and "the tenant has any row" from one uncached
partition read (GetMembershipSnapshotAsync); disabled and application rows
still count as members.
- PreviewWhitelistService caches only an activation; "not activated" is read on
every call, so the activation page leaves at the first poll after approval.
IsApprovedFreshAsync is folded into IsApprovedAsync.
- Removed the dead ClearCacheForTenant, GetMemberRoleAsync and
CanManageBootstrapAsync.
Tests reproduce the sequence over the real caches (auth/me and the first
dashboard request); each part fails against the old caching.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Loader 1.3 downloads into a protected run folder of its own
(%ProgramData%\AutopilotMonitor-Loader-<8 hex>), runs the installer only when it
is signed by the pinned publisher and declares the installer role and version
2.7 or later, removes the folder after a successful run and never touches the
agent folder. Unbound arguments now reach the installer by name; the array
splat bound them by position.
- Installer 2.7 refuses without elevation and decides before it writes: guards
read HKLM through the 64-bit view (a 32-bit host never saw the deployment
marker), guard 5 is covered by the agent check. An agent's folder stays as it
is, a skip without agent removes the folder, an install starts from an emptied,
protected folder and stages the ZIP in a run folder. It moves the MSI's working
directory out, writes stderr only on failure and removes the copy older loaders
left at the fixed path.
- Help texts of both scripts state what customers need; .CHANGELOG, which made
Get-Help ignore the help, gives way to the Platform Changelog.
- OOBE go-script: the installer's shared helpers, its own elevation check
(#Requires is ignored under irm | iex), a run folder.
- Publish: version read line-anchored, the loader's own check on the exact
publish bytes, installers before loaders, -Channel dev (dispatch input).
- Bootstrap MSI 1.2.0; Test-ShouldBootstrapAgent reads the 64-bit view;
Uninstall removes loader run folders and keeps the Deployed marker.
- Tests: Pester unit suites, an integration harness running the real scripts
(loader 1.2 and the MSI 1.0 runner frozen in TestData), BootstrapContractTests;
the gate runs on windows-2022 and windows-2025, once more as SYSTEM, and against
a real standard user.
[skip actions]: published by a dispatch of publish-scripts.yml, channel dev first.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The first reports with context all carried OobeState=completed: Windows'
OOBE flag flips at the desktop, was read at send time after minutes of
retries, and marks the later successful starts too, so it separated
nothing.
Agent: EspPhaseSnapshot reads the ESP registry once with the tracker's
own parser (device_preparation / device_setup / account_setup while
running, <category>_failed, *_done, esp_done, none, unavailable).
DistressDeviceContext (OOBE flag, ESP phase, boot from the uptime
counter) is captured before a registration's first attempt and travels
with the TPM-PSS report; other reports sample it at send time. The
reporter takes a DistressDetail (TPM version, context). sch= is always
set: the events, none, unreadable or timeout. The failed-run record keeps
the ESP phase and OOBE flag, and beside it - never on the wire - the boot
time and the client-certificate thumbprint; the next successful start's
agent_started reports priorRegistrationLastEspPhase, LastOobeState,
RebootedSince (boots more than 2 min apart) and CertChanged, and every
start reports espPhaseAtAgentStart.
Shared/backend: DistressReport.EspPhase (kept as [a-z_] up to 32) and
DeviceBootUtc (kept within 30 days before and 5 minutes after the report),
stored, logged and on the custom event; the TPM panel item carries the ESP
phase. The GA distress detail names the OOBE flag as such and shows ESP
phase and device boot.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A TpmPssUnsupported report classifies the registration start that failed,
not the device: devices with older TPM firmware fail during setup and
connect after it, on 24H2, 25H2 and 26H2 alike.
Agent: every distress report carries the OS build and the Windows setup
state at send time; the TPM-PSS report adds the TPM manufacturer and
firmware and a language-neutral message with the NCrypt codes of the PSS
signs, the native code of the failed handshake and up to two Schannel
System-log events since the registration started (numeric properties
only). The TPM and System-log reads run in parallel for at most 5 s. The
send is awaited, since the run exits right after it. An attempt that got
an answer resets the last exception, so the probe never classifies an
earlier attempt. The run's last error carries the native code. Win32_Tpm
is read by one shared TpmInfoReader (tpm_status unchanged).
Shared/backend: one body cap for both sides (DistressReportMaxBytes 2048);
the agent shortens its message to stay within it. The endpoint keeps OS
build and setup state only in their format, stores the three fields and
adds them to the log and the custom event. GET audit/tpm-pss-unsupported
returns TPM version, OS build, setup state and connectedAt: the last
upload of a certificate-authenticated session of the serial after the
device's last report (bootstrap sessions excluded, six serials per
SessionsIndex query). The bell says the device could not connect and may
connect after setup and a restart.
Web: the Incompatible TPM panel shows TPM and Status columns, uses the
generated types and no longer says "never" or names a Windows version; the
GA distress detail shows the new fields, Cloud PC and the certificate.
Tests: probe message, Schannel tokens, native codes, payload cap, awaited
send, context bound, last-exception reset, connectedAt rule, filter
comparison limit, repository round trip, panel render. A device serial
from production data in a test fixture is replaced.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Tenant Admins add the Push channel; Telegram and e-mail stay Global-Admin-only
- One Push channel per scope on the tenant and ops paths; the dispatcher fans Push out once per send
- The channel also gates session-watch completions and the device test; paired devices stay listed while it is off
- Members whose role comes from an Entra app role pair and receive like table members; a paused device resumes at the owner's next sign-in
- SLA and failure-streak alerts name the tenant; the id moves to a Tenant ID fact
- The legacy single webhook accepts URL webhook providers only
- Portal: Conditional Access hint in the channel editor, neutral role texts
- Landing: push section with the Notify me when done animation and the hero announcement; SignalR is "live updates"
- Privacy, about and FAQ describe push notifications
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- MSAL's in-flight redirect state stays in its tab. storeAuthStateInCookie mirrored it into origin-wide cookies that MSAL reads first, so two tabs signing in at once overwrote each other and the first one failed with state_mismatch, stuck on the landing. A guard test pins the cache config.
- AuthGate keeps the auth-pending overlay for as long as the browser is signed in, so it lifts only once the target route has committed; it used to lift when loading ended and exposed the landing for the whole client navigation. The target's route payload is prefetched while auth/me is in flight.
- A sign-in return that carried a code but ends without an account starts one fresh sign-in from "/", keeping the saved deep link; any further failure in the tab hands over to ProtectedRoute's sign-in card on the deep link, which waits for a click. The auth response is read at module load (the router writes its hydration-time URL back), the tab's temporary MSAL keys are purged before a retry and before the card's sign-in, and loading holds while a redirect leaves the page. New events: auth_redirect_recovered, auth_redirect_recovery_exhausted, auth_redirect_failed (error codes only).
- host.json adds Access-Control-Max-Age: 7200 through the http customHeaders. The Functions host answers every preflight itself and its CORS policy carries no max-age; its custom-headers middleware runs outside the CORS middleware, so the header lands on the preflight too. A guard test pins the entry to exactly that header. The CORS startup log goes: it read settings this host does not use, at a level that never reached the telemetry.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
None of them had a production caller. DelegatedAdminService.GetAssignmentsForUpnAsync and GetAssigneesForTenantAsync, DelegatedAdminService.SetStatusAsync with IAdminRepository.SetDelegatedAdminStatusAsync (only a test called it; the operator revoke is RemoveAsync), and IAdminRepository.GetDelegatedAdminAsync and IsMcpUserAsync, which only had their own implementation.
Revoke_InvalidatesCache now pins the revoke the operator route runs (RemoveAsync); Disable_InvalidatesCache covers SetEnabledAsync.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Removing or disabling a member under Access Management left their assignment to the tenant's self-service group in place, and the delegated scope never re-checked the membership, so the person kept reading every managed customer.
- Scope: an assignment to msp-{owner} confers the group's tenants only while the assignee holds an enabled TenantAdmins row in the owner (inside the cached row resolution, fail-closed, Warning when ignored). Operator groups and direct grants are unchanged.
- Access Management: remove and disable end the self-service assignment (DelegatedGroupAccess DELETE under every managed tenant, reason member-removed / member-disabled) before the existing stream cut. Re-enabling restores nothing.
- Assigning requires an enabled member and re-reads the membership after the write; the Global Admin group route follows the same rule for a self-service group (enabled member of the owner, bound to it, read-only).
- The SignalR group join resolves the delegated scope uncached, so the re-join right after a revoke cannot be admitted by another instance's warm cache.
- Web: Delegated Access offers only enabled members and says that removing or disabling a member ends their access; the Global Admin page hides the role select for self-service groups.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ErrorCodeCatalog.TryLookupDetailed normalised any 0x-prefixed value without
checking its characters and then ran int.Parse on the low word, so a value
such as 0x8007zzzz threw FormatException. The backend enricher runs that
lookup over device-authored event data on every GET /api/sessions/{id}/events,
so one such event answered HTTP 500 for every reader of the session timeline.
- The catalog parses the hex body once with uint.TryParse (hex digits only)
and derives the HRESULT_FROM_WIN32 low word from the number; it now returns
null for any input instead of throwing.
- ErrorCodeEnricher.EnrichEvents isolates each event: a failed lookup leaves
that event unenriched, logs a warning and the page still returns 200.
- Web errorCodeMap.ts gets the same hex-only check: parseInt read
0x8007643z as 1603 and labelled the garbage ERROR_INSTALL_FAILURE.
MCP already validated; a test pins it.
Tests: the reported shapes return null, an exhaustive sweep of short tails
after every prefix never throws, and the enricher keeps the rest of the page
enriched when one event fails. Against the previous catalog 10 of the new
tests fail.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Five platform reads take ?tenantId= only as a filter, but their QueryParam
scoping also admitted a delegated (MSP) reader of the named tenant:
global/session-reports, global/rule-submissions, global/metrics/mcp-usage,
global/metrics/mcp-usage/daily and global/metrics/verdict-calibration. They
carry operator data (admin notes, contact data and reviews, an organization's
accounts, classifier diagnostics), so they are catalogued None now; the
handlers still read ?tenantId= themselves, and Global Admins and Global
Readers keep the filter.
- A catalog ratchet in PolicyEnforcementMiddlewareTests lists every platform
route a delegated caller reaches; a new one fails the build until it is
added on purpose. The TenantScoping.QueryParam doc states the rule.
- The middleware took the cross-tenant decision on the first ?tenantId=
value while the handlers read all values comma-joined. It now decides on
the joined string, with ASP.NET Core's parser (the handlers parse Url.Query,
where System.Uri has already escaped a stray '%').
- The rule pages show a delegated reader the admin-only hint instead of the
submission list, as they already did for an Operator or Viewer.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A rating from the feedback bubble reached only the direct Telegram ping: it was the one customer
submission without an ops event and a bell. It now raises FeedbackRatingReceived (Tenant, Info) and a
Global Admin bell, so an alert rule can route it to the Push channel. Message and details carry the
stars and the comment length, never the comment or the sender; the Telegram ping stays until it is
retired. The answer route admits members only (MemberRead, like feedback/general): it now feeds an
operator alert, and the bubble only ever asks members.
The alert rules page was linked from no menu. Ops > Alert Rules points at it, and a search in its
header narrows the rules by event type and category; it filters the display only, so a save keeps
every rule. navSectionReachability.test.ts holds the sidebar and the section registries together in
both directions. The ops events page and the rules search share one SearchInput.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The device model table on the app detail page put the terminal-only failure
rate (failed / (failed + succeeded)) next to an Installs count that also held
in-progress, incomplete and skipped rows, so 11 failed of 28 read as 50%
without a way to check it.
- Version and device-model rows carry Succeeded on the wire.
- The device model table shows Finished (the rate's base) with a muted
"+N not counted"; the Total card names the same rows.
- The PDF report gets Finished and Not counted columns in both tables and
the hint on Total installs.
- The device-model sample floor counts exactly the rate's base; skipped rows
no longer count toward it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Usage Metrics (tenant page and the platform section) shows the window's
sessions in three groups that each sum to the session count:
Autopilot v1 or Device Preparation, Entra or hybrid joined, and the
Autopilot deployment mode (user-driven, self-deploying, pre-provisioned).
Type and mode follow the fleet enrollment class; user-driven no longer
counts the IsUserDriven flag, which every registration sets. The counts
come from the session rows on each request, so every retained session is
covered without a counter.
- DeploymentTypeMetrics gains the new counts and shares; the usage
projection reads EnrollmentType, IsSelfDeployingProfile, IsHybridJoin
- platform and tenant compute share one builder per block
- Platform Usage takes ?days= like the other analytics pages
- the usage pages share their blocks (components/UsageMetricsBlocks.tsx);
enrollment class labels and the time-attribution segments each live in
one place, the duplication baseline shrinks by seven clones
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A push carries only category, event and severity for an ops alert, so the
portal link is the way to the details, and ops alerts had none: the event
id was minted inside the save and never reached the dispatch.
Backend: SaveOpsEventAsync sets OpsEventEntry.Id (OpsEventId,
{Category}_{RowKey}); WriteAsync hands it to the dispatch, and BuildAlert
adds "Open in the portal" -> /admin/ops/?eventId=<id> on the plain and the
payload path, so every ops channel (push, Telegram, Teams, Slack, Discord,
e-mail, JSON) gets it. GET global/ops-events?id= is a point read (400 for a
malformed id, an empty list for a deleted one). hardware_rejected links to
the hardware whitelist (one path constant, also for the bell); the session
watch link names the session's tenant; TenantSignup's portalPath keeps its
tenant query. Only a session link makes a push session-bound: tag and Topic
no longer key on any "id=" in the link.
Web: /admin/ops?eventId= opens that event's detail from the point read
(matched by id), and the detail links to the tenant and to the event's
related page (same-origin paths only) besides the session, which now
carries the tenant. CopyButton and the cursor pager footer are shared
components instead of four and three copies; the duplicate baseline shrinks
by five entries.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
When the event limit or the platform operator blocks a session's device, the
session now carries a block record: Block* fields on the session and its index
row, a server-authored device_blocked timeline event, an interim analyze run for
that event and, on the auto path, an audit entry. The block list itself stays
platform-only.
- Portal: a card above the failure snapshot says the monitoring was paused
(orange) or ended (red), when, until when and, for the event limit, the
session's event count against the limit. The Blocked pill reads the record for
every role; Global Admin mode still adds the live block list.
- A Kill is device-wide until it expires, so its record carries the end and the
card says new enrollments on the device are not monitored until then.
- Status: the stalled and timeout reasons name the block instead of a silent
agent; the classification is unchanged.
- Rules: ANALYZE-ENRL-006 explains the pause right after it (on_event
device_blocked); ANALYZE-ENRL-002 no longer fires for a blocked session.
- MCP: the coverage block of get_session_summary names the block as a gap.
- The agent-facing block message no longer says "by an administrator".
- BlockCause == event_limit replaces ExcessiveEventsAutoActioned as the
auto-action's idempotency marker.
- Privacy page: the event limit next to rate limiting.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A stall report never overwrote a guard: the ingest branch for session_stalled
runs after the completion and failure chain, and the transition gate let
Stalled through over every status but Pending. An offline batch carrying
session_stalled next to enrollment_complete left a finished enrollment
Stalled until the timeout sweep reclassified it, and a sweep that lost the
race against a completion could do the same.
IsTerminalTransitionAllowed now allows Stalled only over InProgress, on the
normal and the ETag-retry path alike; the separate Pending guard is folded
into it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
GET apps/fleet-context (MemberRead, QueryParam) counts one app over the
tenant's last 7 days: installed, skipped, failed, incomplete, in progress,
collision rows left out, the three most frequent failure codes. It reads
the apps dashboard's partition query, leaves out the asking session and,
with the culprit's app id, another app of the same name; no session point
reads. AppFleetContextService caches the app's rows per tenant, channel,
name and minute for 60 s (single flight).
The classification is the analytics view's: MetricsMath.IsInstalled
replaces six inline copies, AppsAnalyticsHelper.IsSameApp and
TopFailureCodes are shared by analytics, sessions and the fleet context.
The diagnosis page shows the result as a tile in "Around the failure"
only when other enrollments exist (the read fails soft), with "Show
sessions" on the app page where that page shows the viewed tenant. An
expanded finding shows "Fired in N enrollments in the last 14 days"
through useRuleFireStats and RecentFiresLine, both extracted from the
session page, whose cross-tenant check now ignores case.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The diagnosis page is rebuilt on the server headline instead of rule
confidence: a verdict card per status and cause (culprit tiles, ESP
phases, what an unfinished enrollment still waits for), the
circumstances around it, the steps of the finding that explains it,
where the time went, all open findings grouped by role, team notes and
the deduplicated errors and warnings. It reuses the session page's hook
set, refetches the headline on live activity and opens for Failed,
Stalled, AwaitingUser, Incomplete and Succeeded with open findings, also
from read-only and cross-tenant views.
Rule text: lib/ruleText.ts prunes a finding's explanation and steps like
the notifications do (shared cases in tests/fixtures/rule-text-display,
also run by RuleTemplateInterpolator.InterpolateForNotification), then
parses the markdown and fills values into text nodes only. The session
page's analysis section, the rule card and the dry run render through it;
components/findings/EvidenceChips replaces the two evidence-link copies
and covers correlation pairs.
Vocabularies: DiagnosisPrerequisites names the engine's completion
prerequisites (manifest diagnosis.prerequisites, parity fact against
DecisionEngine.CompletionPrerequisites); the page's texts for every
diagnosis vocabulary are exhaustive over the manifest types.
Fixed on the way: useSessionSignalR held the first render's callbacks,
so ruleResultsReady was lost on an already connected hub; a deep link to
an event dropped its hash before the event loaded; the global analyze
rule update answered 500 for an invalid presentation (now 400); the
rule editor kept the presentation block on create and in JSON mode.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A typed headline built on demand from the session row and every event:
outcome = status; cause from the verdict path (derived for older rows,
flagged as a caveat), and for an explicit failure from the ESP evidence
(failed step, error code, culprit apps, AppX candidate, stuck policy
providers). Plus the ESP phase rollup, the ESP profile, what an unfinished
session still waited for, and observation caveats. Never from rule
confidence: Stalled, AwaitingUser and reconciled sessions get no terminal
rule run.
- App facts come from the session's own app events (the app table has no
intent or targeting); only required apps count, a later install clears
an app, a zero error code is none, and a 0x87d1041c ESP failure ranks
the NotDetected app first (real IME payload shape).
- The event read is strict: a storage error is a 500, never a verdict
built from no evidence. 400 for a non-GUID id, 404 for an unknown session.
- Helpers/EventDataReader replaces three private Data-string copies
(classifier, failure snapshot, time attribution) and the list parsing in
EspBlockingSets; EspBlockingSets.FromLatest replaces two backward walks.
- Six vocabularies in the shared manifest section `diagnosis`; parity facts
pin the full and the minimal shape; route in the policy catalog.
- MCP get_session_summary carries it as `headline` right after coverage,
null when the call fails.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
All 54 shipped analyze rules get role, summary, subject and key facts as
{{signal.key}} templates, each with a minor version bump so the self-seed
picks them up. Values were checked against stored findings, the rows that
receive the block at read time.
- ENRL-001 repeats the outcome only: the 0x87D1041C paragraph and the
placeholder note are gone, labels read "ESP step that failed" / "ESP
error code", remediation is "Find the cause". This changes its alert text.
- ENRL-002 says the sweep "closed" the session (it can end Incomplete).
- ID-002 and ID-004 gain evidence-only conditions (unexpected account
names; IME token failure count and codes); firing and confidence are
unchanged.
- Resolution: facts sharing a label are alternatives, an exact repeat of
label and value is left out (new shared fixture case, web module and
its generated MCP copy).
- Ratchet: every built-in presentation passes the save validator, uses
qualified tokens only, and each token names a condition of the rule and
a key its source records (RuleTemplateInterpolator.TokenRegex internal).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Analyze rules can describe how a finding is shown: a role (outcome, cause, context,
advisory), a one-line summary, the subject it is about and up to six key facts. Like the
explanation, the block is stored raw on the result and resolved where it is rendered.
- Templates: a qualified token {{signal.key}} reads one key of one condition's evidence,
so counts, durations and correlation fields are reachable. RuleEvidenceKeys (Shared,
exported to the manifest) lists the keys each condition source records; tests pin it
against the evaluators.
- Backend: PresentationJson on AnalyzeRules and RuleResults, RuleVersion on results,
ValidatePresentation on create/update, and results stored without a presentation get
the current rule's at read time.
- Web + MCP: lib/rulePresentation.ts resolves the block (unrecorded facts left out, summary
falls back to the explanation). The MCP now runs generated copies of the web modules
instead of a hand-kept mirror (npm run generate:mcp-logic, freshness tests in both
suites); the duplicate baseline shrinks by one.
- MCP output: get_session and get_session_summary return the resolved block (issues gain
ruleId, role, summary, subject, facts); test_analyze_rule previews it; validate_rule
checks qualified tokens against the key catalog and lints the subject.
- Shared case files: rule-template-interpolation gains the qualified-token cases,
rule-presentation is new.
No rule carries a presentation yet; renderers behave as before until rules gain one.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Install-AutopilotMonitor.ps1 v2.6: Get-AgentPayloadRejections (Test-FileCatalog
-Detailed; signature Valid, signer -like the pin, files exactly as catalogued) and
Install-VerifiedAgentPayload (extract into Agent-Install, verify there, then move into
the agent folder; a rejection deletes the staging folder, so the next run starts clean).
Replaces the name-filtered Get-UnverifiedAgentBinaries; the dead "already installed"
branch behind guard 5 is gone.
- OOBE go-script: embeds the installer's marked block verbatim (EmbeddedResource) and
verifies before --install; pin and catalog name come from the Shared constants; the
golden is updated, new tests pin the embedding and the order check-before-install.
- build-agent.yml: the release verify step dot-sources the installer and runs its own
function on the final ZIP.
- Pester: every decision branch, a real unsigned catalog, the staging flow (64/64).
- Privacy page: the catalog claim; LAST_UPDATED.
Merge only after a catalog ZIP is live on stable (plan step V0): pushing this to main
publishes the installer to every loader on its next run.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A caller whose token shows it is not a member of the token's tenant (acct = 1, a #EXT# user name,
or an idp naming another tenant, a personal Microsoft account or an external sign-in provider) no
longer seeds DomainName/OnboardedBy or queues the activation, is never the first-user Tenant Admin,
and cannot set the activation-page address in the roleless signup window. Positive evidence only:
a token without these signals stays a member's, and tid/idp are read under their inbound-mapped
names as well. Activation promotes OnboardedBy only; the UpdatedBy fallback is gone and a stored
#EXT# UPN is refused. Explicit grants (TenantAdmins row, app role) still apply.
The once-per-interval log gate of the client-app binding becomes Helpers/LogThrottle and also
throttles the new warning, which the polling activation page would otherwise repeat.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Every stored history entry starts unread. After a new entry the worker sets the
app badge to the unread count through applyBadge, fire and forget: inside
Safari's declarative push event the setAppBadge promise never settles, so it is
never awaited and never handed to waitUntil (pushSw.test.ts pins it). The
history page marks everything read on every visible load, clears the badge,
closes the notifications whose entries the history holds and shows "New" on
what arrived since the last visit. Unpair clears the badge; the wipe push
leaves it to the next open. The payload repeats mutable inside notification,
where Safari 18.4-26.1 reads it, and carries no app_badge.
Bumps internal (plan, D-350, D-351, concept doc).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
"Notify me when done" now also fires on the first genuine pre-provisioning
seal ("Pre-provisioning finished") on the gate of the Pre-Provisioning
Completed alert, and the watch is consumed there; watching the user phase takes
a new click. A watch placed on a sealed session lives 30 days instead of 48
hours. EventIngestProcessor.WatchOutcomeFor is the one gate, the watch text
comes from PushSessionWatchService.WatchMessage. The portal button re-reads the
watch on every status change and names what it waits for. The Success hint
says it covers a finished pre-provisioning; the receiver labels the
preprovisioning_* alerts.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
.NET's $ also matches before a final newline, so "7c9e6679\n" passed the allow-list that guards
the inbound X-Correlation-ID against CRLF and log forgery and was echoed. Such a value does not
arrive over HTTP, but the gate now holds what its comment promises; a new test case pins it
(red with $).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A security client that inspects TLS strips the device certificate on the way to the backend;
the platform then answers 403 "Client Certificate Required" itself, or a proxy answers with
its own page. The agent counted these like backend rejections: five in a row shut it down,
the scheduled task restarted it at every boot, and only the 48-hour emergency break removed it.
- Origin of every 401/403: only a response that echoes the request's X-Correlation-ID is the
backend's verdict and counts toward the auth shutdown; the stopped-app page still counts;
anything else is foreign (client_certificate_required or foreign_response) and never counts.
- Telemetry: a foreign answer keeps the batch and pauses uploads with a backoff from 30 s up to
15 min, without a timer; collectors and the decision engine keep running, the terminal phase
gets one upload attempt, and the agent still leaves the device when the enrollment ends.
- Distress type ClientCertificateNotReceived (9), once per process, with operation, status,
page title and the issuer of the server certificate; both handlers observe it through a
validation callback that returns exactly sslPolicyErrors == None.
- Registration: an intercepted start retries on at most InterceptedRegistrationRetryBoots
further boots (config channel, default 1, agent clamp 0..5, ConfigVersion 41), then ends
monitoring like the emergency break. The break, this give-up and the config kill signal now
share Program.EndMonitoringAtStart.
- Evidence: telemetry_upload_blocked carries reason, operation, statusCode, pageTitle,
serverCertIssuer and firstBlockedAtUtc; agent_started carries priorRegistration* after failed
starts (a local network fingerprint; only whether it changed leaves the device);
agent_metrics_snapshot counts 2xx responses without the echo.
- Backend: distress type 9, the buffer default, InterceptedRuns on the registration-failure
summary; tests pin the correlation echo on 401/403 and the middleware's place in the pipeline.
- Web: distress labels for TPM Incompatible and Cert Not Received; config report row.
- MCP: session coverage names intercepted uploads and the failed registrations before the start.
D-349
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
GET /api/auth/mcp reports the caller's client-capped member role in the token's tenant
(tenantRole; reported, never a grant). The MCP server registers test_analyze_rule and
test_log_pattern only for a platform role or a Tenant Admin: their routes are
TenantAdminOrGlobalReader, so Operators and Viewers only ever got a 403. validate_rule,
author-rule and the authoring guide hand a draft to a Tenant Admin otherwise, and
investigate-failed-session and debug-session end by drafting a rule for a cause that can
recur. Catalog ratchet lowered for tenant users (51,000) and delegated callers (55,500).
D-348
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The root .editorconfig raises IDE0051, IDE0052 and IDE0060 (non-public
methods) to warning, and Directory.Build.props sets EnforceCodeStyleInBuild,
so the warning freeze turns them into build errors in every project, the net48
agent included.
Cleanup to start at zero, each finding checked for leftover vs. missing
behaviour (none was a feature that silently never ran):
- 38 injected dependencies nobody read: loggers, configuration, and services
whose last read had moved to a live replacement (the IME reseed writes
through the rule repository, the offboarding poison OpsEvent comes from the
handler, the vulnerability gate sits in the correlate handler). Removed with
their constructor parameters and the argument at every construction site.
- unused private members: SelfUpdater.IsNewerVersion, a GatherRuleExecutor
wrapper, test constants and an unused test overload.
RjPhaseCompletedFirstDeployment now replaces the two literal 110s it names.
- unused parameters removed together with their arguments; test fakes bound as
method groups use discards.
- the commented-out backend-URL connectivity check and its plumbing are gone;
the reason the backend stays out remains as a comment.
- DiagnosticsPackageService names the blob in its upload-failure warnings, as
its doc already said.
- stale docs fixed on the way: OnboardedAt no longer claims a maintenance
backfill (shared manifest and wire types regenerated), the vulnerability
functions' policy comments match the catalog, a misplaced summary moved to
its method.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The web app kept helpers in two root folders (lib/ and utils/) and named
route-local helper folders either utils/ or lib/. Shared helpers now live in
lib/, route-only helpers in a lib/ next to their route (app/dashboard/lib,
app/sessions/lib). Generated wire types, manifests, guardrails, error codes and
the lastmod file move to lib/generated/, like the MCP server's src/generated/.
The session-search parity cases sit next to sessionSearchQuery.ts.
Pure move, no behaviour change. Imports, the generators (combine.js,
generate-shared-manifest-types.js, sync-error-codes.js, generate-lastmod.js),
the CI drift paths, the two backend tests and the MCP freshness test that read
these files, and the guard tests' registry path and scan list follow the new
locations.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A health row goes once its channel is no longer in the stored channel list of its scope and it
saw no send for 30 days. The channel lists are read strictly, so a failed read skips the scope
instead of reading as "no channels"; the delete is ETag-conditional, so a row a send just
rewrote survives. Runs on the 2-hourly maintenance timer next to the push retention.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Every non-Push send, real or test, is recorded in a NotificationChannelHealth row per channel
(own table, server-side status rule, destination fingerprint resets the history). The settings
show a status dot per channel: Operating normally, Failure rate x%, Error with the cause, or No
deliveries yet. Two failed deliveries in a row raise one bell for the tenant admins (or the GA
bell for ops channels) per failing episode; a failing channel keeps receiving deliveries.
Each transport now has one send path that reports its outcome (WebhookTestResult is
NotificationSendResult, the dispatcher's test entry is SendTestAsync); failures are logged with
tenant and channel id. New routes GET config/{tenantId}/notification-channel-health
(TenantAdminOrGA) and GET global/config/ops-channel-health (GlobalAdminOnly).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- TableStorageService.GetEnrolledTenantIdsCachedAsync: configured tenants with a PlatformStats counter row (RowKey 'current', TotalEnrollments > 0), 60 s single-flight; fail-open to the configured list with a Warning when the counter scan fails
- Session list pages, cross-tenant stats drain, geo window and app-summary snapshot fan out over that subset (353 -> 127 partition queries per page live); fresh-install fallbacks key on the configured list, a configured set nobody enrolled in is an empty page; audit logs, QuerySessionsByDateRangeAsync and maintenance keep the full list
- Web: cross-tenant follow-up pages fetch max(pageSize, 100) (resolveSessionsFetchPageSize); first page and own-tenant list unchanged
- Tests: CrossTenantFanOutTests (enrolled subset, counter 0 / no row, fail-open, unconfigured counter row, empty enrolled set never the legacy scan, bound after the filter, geo + apps), DelegatedBoundedAggregateTests harness, sessionsPageSize.test.ts
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
A session without a Sessions row can never be reached by the cascade, so every row written for it was an orphan by construction. Three layers now keep the platform clean:
- Prevent: IngestTelemetry answers a batch for an unknown session with 410 (same status and body as the cascade lock; request-row marker IngestRefusal = session_unknown | cascade_locked). The derived writers - vulnerability correlation, analyze, manual rescan - stop on SessionDeletionGuard.GetLivenessAsync != Live. The agent treats every 410 as terminal: one TelemetryRejectedSessionUnknown report (exempt from the reporter cooldown), then terminate_session with forced self-destruct (origin session_gone, no final drain, no diagnostics upload); spool items of a foreign session id are dropped at startup and never terminate the current session.
- Detect: OrphanSessionSweepFunction (every 4h at :45, own lease) diffs the Sessions keys against the EventSessionIndex handles - two range drains instead of ~16.8k Sessions point reads per 2h maintenance run - and re-checks every candidate (handle unchanged, row still absent, tombstone) moments before deleting. The in-maintenance sweep raced the retention cascade on their shared 00:00/12:00 tick and reported false orphans; it is gone.
- Clean: every table of the cascade manifest minus the tombstone pair (ratchet in TableLifecycleBucketTests), the inventory counters with the stamp on the handle (claim, then decrement), the handle last with If-Match. A weekly read-only reconciliation of EventTypeIndex against the Sessions keys (Sunday mornings UTC, and on every manual maintenance run) reports residue.
MaintenanceCompleted now carries stepsMs. MCP search_sessions_by_event auto-exhausts empty-but-continuable pages. Manifests regenerated for the new ops event types and AdminConfiguration.OrphanDeepCheckLastRunUtc.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Customer domains, tenant-ID fragments, two full customer tenant GUIDs and
four real session GUIDs had found their way into code comments, test
constants and a fixture comment. They are replaced by neutral descriptions
("field case", "kiosk tenant") and synthetic values; no behaviour changes.
A check of every onboarded tenant's domain and ID prefix against the tracked
files finds no customer identifier left.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- The retention fanout reads the tenant configuration strictly. A storage
failure used to yield the synthesized default (90 days), which deletes
sessions and prunes rows a Pro-365 or never-delete tenant keeps. Now the
tenant is skipped and counted: tenantsFailed on
SessionDeletionMaintenanceCompleted, which turns Warning. A row that
vanished since the tenant listing is skipped too, instead of being
recreated with defaults.
- GetDeviceHistoryAsync throws on a read failure instead of returning "no
history". Both writers merge into the chain they read and replace the row,
so the old behaviour dropped every older ref. The sweep skips such devices
and leaves the FTR aggregates untouched for that tick.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- TableKeySanitizer replaces ten private and inline copies of the key rule
(/, \, #, ? become _) and also removes control characters, which Table
Storage rejects in keys and the HTTP layer rejects in a point read's URL.
Keys without control characters are unchanged; a source ratchet forbids new
copies. The dead UpsertSoftwareInventoryAsync is removed.
- VulnerabilityInventoryLoader hands out inventory text without control
characters, so CPE lookup, report text, side-row identity and counter keys
all see the same text.
- A correlation run whose SoftwareInventory counter writes or contributions
side-row fail writes one SoftwareInventoryCounterWriteFailed ops event
(Maintenance, Warning; counts, error type and status, up to five sample
keys) instead of a per-key entry in the tenant audit trail. The counter work
is awaited inside the invocation instead of a detached Task.Run.
- The new type is listed in the alert-rule catalog; shared manifest and MCP
vocabulary regenerated.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
First-time-right, time-attribution and rule-statistics rows lived on fixed lifetimes (180,
180 and 90 days), a storage bound reused from the never-pruned tables, never weighed against
tenant retention: a Community tenant could read first-time-right rows for days 91-180 through
the API and MCP, a Pro tenant keeping 365 days lost them after 180 and its rule statistics
after 90.
The retention fanout now prunes per tenant, before its session scan: these rows at the
tenant's effective retention (downgrade grace included), the audit trail at
max(180, retention) up to 365. The maintenance cleanups become outer bounds and platform
lifetimes in DataLifetimes: no tenant row beyond 365 days, global first-time-right rows 365,
time attribution 180, global and legacy rule statistics 90, the platform audit partition and
legacy audit rows 180. Device histories already follow their sessions and stay as they are.
The device-journeys window goes to 365 days (route cap and MCP get_device_history). The
SessionDeletionMaintenanceCompleted ops event carries tenantRowsPruned.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A tenant that is Pro only through its managing organization ("Pro (MSP)") could not save a
data retention above 90 days: the patch and revert paths validated against the raw table row,
which never carries the read-time projection, so the edition resolved to Community while the
feature flags and the settings slider offered 365.
The validator now gets a projected copy of the stored row. The row itself stays
projection-free, because the exactly-these-fields verify diffs against it; a projected row
would read as drift and roll every save back.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The host decodes %2F to a path separator before routing, so apps/{appName}/analytics|sessions and
their global twins answered 404 for every app whose name contains '/', before any function ran;
a name with a literal escape such as '%20' was decoded a second time. The routes are now
apps/analytics, apps/sessions, global/apps/analytics and global/apps/sessions with ?appName=, and
the unblock route is DELETE devices/block?serialNumber=&tenantId= (serials may contain '/' too).
Function names, policies and tenant scoping stay; req.Query decodes exactly once. Tests run names
with '/', '%', '+', a space and a literal escape through the catalog and the handlers, and the web
builders put the name into the query. Manifest and wire types regenerated (doc text only).
Rollout: backend, then web right after; an app-detail tab on the old bundle 404s until reloaded.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A receiver holds one pairing, so a Global Admin whose phone is paired in the platform scope
could not get their own tenant's enrollment alerts without replacing that pairing. The push
transport now resolves a tenant channel's recipients as the tenant's Active devices plus the
Active platform devices whose OwnerHomeTenantId is that tenant, filtered by the platform
eligibility rule and deduplicated by endpoint; no other tenant reaches a platform device, there
is no opt-in and no setting. The home tenant's device list shows such a row with
PushDeviceDto.Scope = "platform" (manifest and wire types regenerated); the panel marks it with
a Platform badge and no actions — it is managed under the platform's device list, and the
tenant routes do not know the id. Tests: fan-out incl. a foreign home tenant, a former Global
Admin and a shared endpoint; the list in the home tenant, the owner's and an operator's view,
and the refusing tenant routes.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Every session-watch push carried the tag "session_watch" (SystemMessage set Tag = type), so a
browser that honours tags — Edge on the Windows device, iOS for notifications the worker shows
itself — replaced an earlier completion with the next one; two watched sessions finishing
minutes apart (18:38 and 18:40 UTC today) would have shown as one. The tag is now the payload
id: a test, a pairing or a session watch never replaces another notification. Alerts keep
their per-session tag, and the history is keyed by id and keeps every push either way.
The iPhone kept running the worker without the navigate fix through two opens: a home-screen
app that is only resumed never navigates, and the browser checks for a new worker on
navigation. reconcileDevice now calls registration.update() on every open (fire-and-forget).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
- Get started keeps its tick as the acceptance: the CTA hands ?signupConsent=1 to portal (kept per
tab for 30 minutes, removed from the address bar at once), the first auth/me forwards it, and the
onboarding write records DpaVersion together with DpaAcceptedBy (first-login UPN) and
DpaAcceptedAt.
- Any other first sign-in (the navbar "Sign in") sets DpaAcceptancePending on the new row instead.
auth/me carries the flag; ProtectedRoute renders a blocking dialog with the same sentence and
links as Get started (shared TermsDpaAgreementText) instead of the page; POST
auth/dpa-acceptance writes the acceptance for the caller's own tenant only, a no-op 204 when
nothing is pending.
- Tenants onboarded before this carry no flag and are never asked. Activation is unchanged; the
dialog appears after it.
- The three new fields are system-owned like OnboardedBy (patch deny-list, revert-protected) and
projected by config/all, the GA tenant report and the MCP tenant list.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Both mails were landing in Junk at some recipients although the domain's SPF, DKIM and DMARC
are in order, so the copy loses its marketing markers: the welcome subject names the tenant
("Your Autopilot Monitor access for {domain} is ready", no exclamation mark; the editor shows
it with the domain placeholder), the heading drops its exclamation mark, "Great news" and
"incredibly valuable ... don't hesitate" give way to plain sentences, the closing line is a
thank-you, and both templates keep GitHub as their only feedback link (the personal LinkedIn
link, a second foreign link domain, is gone). User go for the customer-visible wording.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Both template mails also land in Junk at some recipients. The templates are already image-free
and inline-styled, so the remaining technical signals were the missing text/plain part (the
provider derived one) and the missing automation header. HtmlToText renders the effective
template (default or operator override) as the text part — head, style, script and comments
dropped, blocks and list items as lines, links with their URL, entities decoded — and
EmailService sends every mail with Auto-Submitted: auto-generated / X-Auto-Response-Suppress,
shared with the alert mails. Wording changes to the customer-visible templates are deliberately
not part of this commit.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The first channel mail landed in Junk although SPF, DKIM (mte1/mte2) and DMARC are in order,
so the content was the tell: a bracketed, shouting subject built from the machine title
"Ops Alert: Platform/PushDeliveryFailed", a raw JSON block, hardly any prose, no text part.
Now: subject "Autopilot Monitor: Push delivery failed (Platform)" (event name as words, no
tags, no severity word), one introductory sentence naming the severity and category, the
event as a heading, facts and the opted-in payload as key/value tables (nested values as
compact JSON, non-object payloads still capped), a visible portal link, a footer that says
why the mail was sent, a plain-text part of our own (auto_text off for alerts) and the
RFC 3834 Auto-Submitted header. Template mails are unchanged.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
A test mail that Mandrill accepted did not arrive; the backend had logged nothing beyond
"accepted". SendViaMandrillAsync now returns the status (sent/queued/rejected) and the message
id, the channel test reports them ("Provider status: queued, id ..."), and the EmailDelivery
custom event carries them as status and providerIds — enough to look a mail up in the
provider's activity log.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
WebhookProviderType.Email = 60 joins Telegram and Push behind the Global-Admin gate: it sends
under the platform's sender identity and provider key (the welcome/farewell path,
noreply@autopilotmonitor.com). The channel's Url names one to five plain recipient addresses
(ValidateEmailDestination, no SSRF URL check). EmailNotificationService renders the alert
(EmailAlertRenderer: severity-prefixed subject, escaped HTML with summary, facts, sections,
links, and the structured payload as preformatted text when the rule opted in) and sends one
message per recipient through EmailService.SendMessageAsync; outcomes land in the EmailDelivery
custom event and the log, never in an ops event (a failing e-mail channel must not feed itself).
The dispatcher takes the transport through a chained fourth-parameter constructor so the
existing mocks keep working. ChannelEditor offers the provider to Global Admins with a
"Recipients" field. Tests: routing, validation, gate, renderer, send against a provider stub.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
TenantSignup, SessionReportSubmitted, DiagFilesReportSubmitted and RuleSubmissionReceived
(category Tenant, Info) are recorded where the backend pings the Telegram chat directly, so an
alert rule can route them to any channel — the Push channel included. Like FeedbackReceived,
message and details carry ids and a portal path only, never the submitter or a comment (the
user-id column keeps the submitter for the admin view and never reaches a channel). The ops
event is awaited before the response; the Telegram pings stay untouched until they are retired.
Dual-registered (constants, All, web rule list), vocabularies regenerated, health marker row.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Oliver Kieselbach