Six adversarial-review findings, each verified against the code first.
F1 (HIGH) CLI HttpClient capped every call at min(30s, caller timeout),
silently truncating deploy site's 5-minute BulkDeployTimeout and the
5-minute bundle export/preview/import calls — which printed a fake
"504 Request timed out" while the server kept working. HttpClient.Timeout
is now Timeout.InfiniteTimeSpan (the per-call CTS is the single overall
deadline, connect included) with the connect phase bounded separately on
SocketsHttpHandler.ConnectTimeout. The env override is renamed to
SCADABRIDGE_HTTP_CONNECT_TIMEOUT_SECONDS to match its new meaning.
F2 (HIGH) StaleInstanceProbe's process-static memo served stale hashes
because nothing bumped the watermark on three paths:
(a) BundleImporter commits through the raw DbContext, so no import ever
moved the watermark — a second import overwriting the same template
could be OMITTED from ImportResult.StaleInstanceIds. It now bumps
once per apply ATTEMPT: after the commit, and after the rollback too
(the probe runs pre-commit, so a rolled-back attempt leaves memos for
state that never landed; bumping on both paths is the simplest
correct shape, versus threading transaction awareness through a
process-static cache).
(b) CollectWatermarkBumps' default: arm silently no-op'd, contradicting
its own doc. It now sets unattributed=true — an over-broad bump costs
extra work, a missed one produces stale work.
(c) DataConnection edits route through SiteRepository.SaveChangesAsync,
which had no watermark at all, yet Protocol/Primary+Backup config/
FailoverRetryCount are revision-hash inputs. It now bumps (BumpAll —
a connection has no owning template) after a commit that touched one.
F3 (MED) CLI TemplateTableProjection read child ARRAYS, but ListTemplates
now returns database-projected TemplateSummary rows, so template list
printed all zeros. It now prefers the *Count scalars and falls back to
array length (template get still returns full entities). --detail help
text and README corrected: a listing cannot yield definitions, so --detail
renders the raw summary payload and template get --id is the full dump.
F4 (MED) DeploySiteAsync staged every PendingDeployment in phase 1 against
a 5-min TTL while phase 2 reached them one batch at a time, so tail
instances' fetch tokens could expire before their command was sent.
Staging moved into phase 2, immediately before each send; prepare keeps
its flatten/validate/record work. The staging write is the phase's only
repository touch and is serialised behind a 1-permit semaphore, so the
non-thread-safe DbContext constraint holds and the sends stay concurrent.
F5 (MED, latent) DeploySiteAsync leaked every held operation lock if
cancelled — a wedged per-instance semaphore is permanent for the process.
Phase 2 no longer throws (cancellation is recorded as a per-instance
outcome so phase 3 still runs), and an escape from phase 1 or 3 now
unwinds every unfinalised entry: Failed status + lock release.
F6 (LOW) ScriptCompileVerdictCache's promotion wrote hot directly,
bypassing SegmentCapacity (true ceiling 3x against a documented 2x).
Promotion now goes through Store, keeping generational semantics; _hot
and _cold are volatile.
Tests: CLI 396, DeploymentManager 133, ManagementService 494,
TemplateEngine 478, ScriptAnalysis 60, Transport 157, Transport
integration 106, ConfigurationDatabase 366 — all green, 0 build warnings.
The F2/F4/F5 regression tests were each confirmed to FAIL with their fix
reverted.
#05-T24. Three low-severity cleanups, each TDD'd:
- TransportOptions.MaxConcurrentImportSessions (8): BundleSessionStore.Open rejects
a new session past the cap (evicting expired entries first), bounding the N×~200 MB
decrypted-content footprint. Soft cap (count/add not atomic).
- ScriptCompiler.TryCompile now joins ALL forbidden-API violations / compile errors
(string.Join instead of [0]) so an operator sees every problem at once; the T15
verdict cache stores the joined string.
- SemanticValidator emits an advisory warning when a CallScript target resolves
ONLY via the composed leaf-name fallback (dynamic child path, not statically
verified) instead of silently accepting it.
Claude-Session: https://claude.ai/code/session_01MtdgwpEeCUn6cUA5f1LMPj
Remediation from the full per-module code review at 4307c381 (findings recorded
separately in code-reviews/).
Highs fixed:
- DeploymentManager-025/SiteRuntime-031: stop broadcasting notification lists + SMTP
configs (incl. credentials) to sites; site purges already-persisted rows on apply
(enforces the central-only delivery design; clears plaintext SMTP creds at rest).
- DataConnectionLayer-023: guard the native-alarm subscribe path against the
mid-flight-unsubscribe adapter-feed leak (mirrors the DCL-021 tag-path fix).
- SiteEventLogging-024: normalize From/To query bounds to UTC (the -016 fix the
audit trail claimed but never committed).
- KpiHistory-001: add an in-flight guard to the recorder sample tick.
- ScriptAnalysis-001: harden the trust analyzer's TPA-absent fallback (resolve
forbidden anchors in the minimal reference set; warn on degraded mode) — anchors
added to validation references only, never the compile gate.
(InboundAPI-026 left to the feat/ipsen-movein effort per owner decision.)
Medium/Low: DM-026 deterministic deploy-status tiebreaker; SR-027/028/029/030
native-alarm leak/phantom-active/delete-during-redeploy fixes; AL-013/014/016;
TE-024 (folder-mutation audit rows now persisted)/025; SF-025 gauge-provider
clear-on-stop; ESG-025/026; SEC-023/024/025; SCA-007/008/009; plus doc/test
accuracy COM-023/024, HOST-025/026, HM-024/025, NS-027/028.
Full-solution build 0 warnings; ~3560 tests across 18 touched suites green.
Add code comments in ValidateConnectionBindingCompleteness explaining
that the unbound-attribute branch also covers the silently-dropped
stale-binding case (cross-reference FlatteningService.ApplyConnectionBindings),
and that the `continue` skips the exists-at-site check for unbound attrs.
Add two new tests:
- FlatteningPipelineConnectionBindingTests: stale DataConnectionId (999)
not present in site connections → flattener drops it silently →
validator reports ConnectionBinding Error, IsValid false.
- ValidationServiceTests: enforce:true + siteConnectionNames:null on a
properly-bound attribute → no ConnectionBinding error (exists-at-site
check stays inert when site set is not supplied).
Pre-deployment validation only WARNED when a data-sourced attribute had no
connection binding, so an instance with unresolved bindings still passed IsValid
and could deploy. There was also no check that a binding resolves to a connection
that actually exists at the target site.
- ValidationService.Validate gains an opt-in `enforceConnectionBindings` flag
(default false) plus a `siteConnectionNames` set. Default-false keeps the
template DESIGN-TIME path (ManagementActor.HandleValidateTemplate) non-blocking,
since bindings are legitimately set later at instance/deploy time. The DEPLOY
path (FlatteningPipeline) opts in (true) so:
* a data-sourced attribute with no binding is now a deploy-gating Error;
* a binding to a connection that does not exist on the target site is an Error.
Static (non-data-sourced) attributes are never flagged.
- FlatteningPipeline computes the site-connection-names set from the loaded site
data connections (mirroring M2.1's alarmCapableConnectionNames) and threads it in.
- Tests: TemplateEngine.Tests covers design-time warning / deploy-time error /
static-ok / exists-at-site / non-existent-connection. New
FlatteningPipelineConnectionBindingTests proves the deploy path enforces it.
Mark M2.7 + M2.8 completed in the plan task tracker.
SplitCallArguments now skips C# line (`//`) and block (`/* */`) comments when
tokenizing the argument list, so a comma inside a comment no longer produces a
spurious arg-count mismatch. IsNumericLiteral now explicitly rejects tokens
whose first non-sign character is `_` or a letter (e.g. `_2`), and restricts
underscore digit-separators to positions after at least one digit, preventing
identifier-shaped tokens from being inferred as Integer/Float.
#20 return-type: when a CallScript/CallShared result is assigned directly into
a typed local declaration (optionally awaited, optionally via an Instance./
Scripts./Parent./Children["x"]. receiver), compare the LHS declared type
against the target script's declared ReturnDefinition and flag clear
cross-category mismatches (ReturnTypeMismatch). Previously BuildReturnMap was
built but never read.
#21 argument-type: positional call arguments are now split (paren/brace/bracket
+ string-literal aware) and each literal-inferable argument is checked against
the target's declared parameter type (ParameterMismatch), not just the count.
Conservative — only CLEAR primitive mismatches (String/Integer/Float/Boolean)
are flagged; Integer<->Float widening is tolerated. Unknown/Object/List
declarations, var/untyped/unused/expression-embedded assignments, and
non-literal arguments (variables, member access, method/await chains, casts,
object/array initializers, compound or concatenated expressions, interpolated
strings) are never flagged. Inference limits documented in code.
Adds 16 SemanticValidatorTests covering mismatch detection, correct-call pass,
and the dynamic/unknown no-false-positive cases.
Joseph Doherty