engineering-platform 2.3.67__tar.gz → 2.3.70__tar.gz

{engineering_platform-2.3.67 → engineering_platform-2.3.70}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: engineering-platform
-Version: 2.3.67
+Version: 2.3.70
 Summary: Local-first Engineering Platform execution operations runtime
 Author: Peter van Tol
 License: MIT

{engineering_platform-2.3.67 → engineering_platform-2.3.70}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "engineering-platform"
-version = "2.3.67"
+version = "2.3.70"
 description = "Local-first Engineering Platform execution operations runtime"
 readme = "README.md"
 requires-python = ">=3.14,<3.15"

{engineering_platform-2.3.67 → engineering_platform-2.3.70}/src/engineering_platform/ENGINEERING_PLATFORM_CONFIG.json

@@ -3,7 +3,7 @@
   "platform": {
     "id": "engineering-platform",
     "name": "Engineering Platform",
-    "version": "2.3.67",
+    "version": "2.3.70",
     "generation": 2,
     "documentation_namespace": "engineering-platform",
     "capability_registry_version": 1

{engineering_platform-2.3.67 → engineering_platform-2.3.70}/src/engineering_platform/ENGINEERING_PLATFORM_VERSION.json

@@ -1,15 +1,15 @@
 {
   "bootstrap_contract": "2026.12",
   "checkpoint_format": 1,
-  "dashboard_version": "2.3.67",
+  "dashboard_version": "2.3.70",
   "handoff_protocol": 1,
   "memory_format": 2,
   "minimum_codex_cli": "0.146.0",
   "inbox_protocol": 1,
-  "platform_version": "2.3.67",
+  "platform_version": "2.3.70",
   "report_format": 2,
-  "runner_version": "2.3.67",
+  "runner_version": "2.3.70",
   "status_model": 1,
   "storage_schema": 44,
-  "watcher_version": "2.3.67"
+  "watcher_version": "2.3.70"
 }

{engineering_platform-2.3.67 → engineering_platform-2.3.70}/src/engineering_platform/assets/dashboard.js

@@ -349,10 +349,22 @@ function phaseAwareRange(estimate) {
   if (estimate?.phase_aware !== true || samples < 2 || !Number.isFinite(lower) || !Number.isFinite(upper)) return null;
   return [Math.max(1, Math.round(lower / 60)), Math.max(1, Math.ceil(upper / 60))];
 }
+function exceededEstimateContext(durationEstimate, elapsedContext = "") {
+  const values = [elapsedContext, t("estimate.exceeded_context")];
+  if (hasHistoricalEstimate(durationEstimate)) values.push(historicalContext(durationEstimate, ""));
+  return values.filter(Boolean).join("\n");
+}
 function estimate(x, durationEstimate = {}) {
   const phase = x.current_phase || "";
   const phaseRange = phaseAwareRange(durationEstimate);
   if (["INITIALIZE", "EXECUTE_AGENT", "REPAIR_AGENT", "FINALIZATION_REPAIR_AGENT", "FINALIZE_AGENT", "REPOSITORY_CLEANUP"].includes(phase) && phaseRange) {
+    const remainingUpperSeconds = Number(durationEstimate?.remaining_upper_seconds);
+    if (Number.isFinite(remainingUpperSeconds) && remainingUpperSeconds <= 0) {
+      return {
+        summary: t("estimate.exceeded"),
+        context: exceededEstimateContext(durationEstimate),
+      };
+    }
     const [minimum, maximum] = phaseRange;
     return {
       summary: t("estimate.remaining", { minimum, maximum }),
@@ -375,6 +387,12 @@ function estimate(x, durationEstimate = {}) {
       remainingMinimum = Math.max(1, minimum - elapsed),
       remainingMaximum = Math.max(remainingMinimum, maximum - elapsed);
     const elapsedContext = t("estimate.elapsed", { elapsed, minutes: pluralMinutes(elapsed) });
+    if (maximum - elapsed <= 0) {
+      return {
+        summary: t("estimate.exceeded"),
+        context: exceededEstimateContext(durationEstimate, elapsedContext),
+      };
+    }
     return {
       summary: t("estimate.remaining", { minimum: remainingMinimum, maximum: remainingMaximum }),
       context: hasHistoricalEstimate(durationEstimate)

{engineering_platform-2.3.67 → engineering_platform-2.3.70}/src/engineering_platform/assets/dashboard_locales.mjs

@@ -251,6 +251,8 @@ export const DASHBOARD_MESSAGES = {
     "estimate.cleanup": "Cleanup in progress",
     "estimate.cleanup_context": "Remaining time depends on the local repository.",
     "estimate.elapsed": "{elapsed} {minutes} elapsed.\nBased on execution size, phase and elapsed time. No live Codex progress or token usage.",
+    "estimate.exceeded": "Estimate exceeded; remaining time uncertain",
+    "estimate.exceeded_context": "The prior range has elapsed; no reliable remaining time is available.",
     "estimate.finalizing": "Finalization in progress",
     "estimate.finalizing_context": "Remaining time is only reliable with live Codex progress.",
     "estimate.historical_context": "Refined with {count} comparable completed executions using the same reported runtime settings.",
@@ -899,6 +901,8 @@ export const DASHBOARD_MESSAGES = {
     "estimate.cleanup": "Opschoning in uitvoering",
     "estimate.cleanup_context": "De resterende tijd hangt af van de lokale repository.",
     "estimate.elapsed": "{elapsed} {minutes} verstreken.\nGebaseerd op opdrachtomvang, fase en verstreken tijd. Geen live Codex-voortgang of tokenverbruik.",
+    "estimate.exceeded": "Schatting overschreden; resterende tijd onzeker",
+    "estimate.exceeded_context": "De eerdere bandbreedte is verstreken; er is geen betrouwbare resterende tijd beschikbaar.",
     "estimate.finalizing": "Finalisatie in uitvoering",
     "estimate.finalizing_context": "De resterende tijd is pas betrouwbaar met live Codex-voortgang.",
     "estimate.historical_context": "Verfijnd met {count} vergelijkbare voltooide uitvoeringen met dezelfde gerapporteerde runtime-instellingen.",
@@ -1547,6 +1551,8 @@ export const DASHBOARD_MESSAGES = {
     "estimate.cleanup": "Bereinigung läuft",
     "estimate.cleanup_context": "Die Restzeit hängt vom lokalen Repository ab.",
     "estimate.elapsed": "{elapsed} {minutes} vergangen.\nBasierend auf Auftragsumfang, Phase und verstrichener Zeit. Kein Live-Codex-Fortschritt oder Tokenverbrauch.",
+    "estimate.exceeded": "Schätzung überschritten; Restzeit ungewiss",
+    "estimate.exceeded_context": "Die bisherige Zeitspanne ist verstrichen; es ist keine verlässliche Restzeit verfügbar.",
     "estimate.finalizing": "Finalisierung läuft",
     "estimate.finalizing_context": "Die Restzeit ist nur mit Live-Codex-Fortschritt zuverlässig.",
     "estimate.historical_context": "Verfeinert mit {count} vergleichbaren abgeschlossenen Ausführungen mit denselben gemeldeten Laufzeiteinstellungen.",
@@ -2129,6 +2135,8 @@ export const DASHBOARD_MESSAGES = {
     "estimate.cleanup": "Nettoyage en cours",
     "estimate.cleanup_context": "Le temps restant dépend du dépôt local.",
     "estimate.elapsed": "{elapsed} {minutes} écoulées.\nBasé sur la périmètre de l’exécution, la phase et le temps écoulé. Aucune progression Codex en direct ni consommation de jetons.",
+    "estimate.exceeded": "Estimation dépassée ; temps restant incertain",
+    "estimate.exceeded_context": "La plage précédente est écoulée ; aucun temps restant fiable n’est disponible.",
     "estimate.finalizing": "Finalisation en cours",
     "estimate.finalizing_context": "Le temps restant n’est fiable qu’avec la progression Codex en direct.",
     "estimate.historical_context": "Affiné à partir de {count} exécutions terminées comparables utilisant les mêmes paramètres d’exécution signalés.",
@@ -2711,6 +2719,8 @@ export const DASHBOARD_MESSAGES = {
     "estimate.cleanup": "Limpieza en curso",
     "estimate.cleanup_context": "El tiempo restante depende del repositorio local.",
     "estimate.elapsed": "{elapsed} {minutes} transcurridos.\nBasado en el alcance de la ejecución, la fase y el tiempo transcurrido. Sin progreso en directo de Codex ni consumo de tokens.",
+    "estimate.exceeded": "Estimación superada; tiempo restante incierto",
+    "estimate.exceeded_context": "El intervalo anterior ya transcurrió; no hay un tiempo restante fiable.",
     "estimate.finalizing": "Finalización en curso",
     "estimate.finalizing_context": "El tiempo restante solo es fiable con progreso de Codex en directo.",
     "estimate.historical_context": "Afinado con {count} ejecuciones completadas comparables que usan la misma configuración de ejecución informada.",

{engineering_platform-2.3.67 → engineering_platform-2.3.70}/src/engineering_platform/execution_executor.py

@@ -26,7 +26,7 @@ from .codex_observability import codex_final_message as _codex_final_message, ex
 from .evidence_projection import ToolProxyEnvironment
 from .execution_context import additional_workspace_write_roots
 from .execution_errors import CodexHandoffTimeout, CodexInvocationError, RunnerError
-from .execution_timeout_policy import SPECIALIST_REVIEW
+from .execution_timeout_policy import AUTONOMOUS_QUALITY_CONTROL, SPECIALIST_REVIEW
 from .execution_models import AgentResult
 from .platform_version import detected_codex_cli_version
 from .provider_usage import churn_from_jsonl, usage_from_jsonl, usage_snapshots_from_jsonl
@@ -44,7 +44,7 @@ _CODEX_USAGE_LIMIT = re.compile(
 )
 _QUALITY_EVIDENCE_ACTIVITIES = frozenset({"REFACTOR", "TEST_COVERAGE", "DOCUMENTATION", "VALIDATION", "NO_CHANGE_REQUIRED"})
 MAX_RETAINED_VALIDATION_OUTPUT_CHARACTERS = 8_000
-REVIEWER_INVOCATION_TIMEOUT_SECONDS = SPECIALIST_REVIEW.seconds
+MANDATORY_ASSURANCE_REVIEWERS = frozenset({"quality", "security"})
 _VALIDATION_STREAM_LIMIT = MAX_RETAINED_VALIDATION_OUTPUT_CHARACTERS // 2
 _UNITTEST_FAILURE = re.compile(r"^(?:FAIL|ERROR): [^(]+ \(([^)]+)\)$", re.MULTILINE)
 _UNITTEST_COUNTS = re.compile(r"FAILED \((?P<details>[^)]*)\)")
@@ -53,6 +53,13 @@ _TURN_ABORTED = re.compile(r'"type"\s*:\s*"turn_aborted"[^\n]*"reason"\s*:\s*"in
 _ASSESSMENT_STATES = frozenset({"COMPLETE", "WAITING", "FAILED", "BLOCKED"})
 
 
+def _reviewer_invocation_timeout_seconds(selection: ReviewerSelection) -> int:
+    """Apply the autonomous assurance boundary only to mandatory reviewers."""
+    if selection.reviewer in MANDATORY_ASSURANCE_REVIEWERS:
+        return AUTONOMOUS_QUALITY_CONTROL.seconds
+    return SPECIALIST_REVIEW.seconds
+
+
 def _parse_validation_assessment(raw: object) -> AgentResult:
     """Validate the small read-only contract before projecting AgentResult.
 
@@ -490,7 +497,7 @@ class CodexCliClient:
                     "--output-schema",
                     str(schema_path),
                         reviewer_prompt(selection, objective, evidence),
-                    ), environment=environment, timeout=REVIEWER_INVOCATION_TIMEOUT_SECONDS,
+                    ), environment=environment, timeout=_reviewer_invocation_timeout_seconds(selection),
                 )
             self.last_context_escalations = proxy.context_escalations()
             self.last_execution_seconds = round(time.monotonic() - started, 3)

{engineering_platform-2.3.67 → engineering_platform-2.3.70}/src/engineering_platform/execution_host.py

@@ -1418,12 +1418,20 @@ class EngineeringRunner:
         """Return whether an agent supplied bounded evidence of a failed local check."""
         if not result.validation_evidence:
             return False
-        summaries = " ".join(
-            item.get("result", "").casefold()
-            for item in result.validation_evidence
-            if isinstance(item, dict)
-        )
-        return any(token in summaries for token in ("fail", "failed", "timeout", "timed out", "error"))
+        for item in result.validation_evidence:
+            if not isinstance(item, dict):
+                continue
+            summary = item.get("result", "").strip().casefold()
+            # Result summaries are dispositions followed by bounded detail.
+            # A passing disposition takes precedence over words such as
+            # "fail-closed" or "failure mode" in that descriptive detail.
+            if re.match(r"^(?:pass(?:ed)?|success(?:ful|fully)?|succeeded)\b", summary):
+                continue
+            if re.match(r"^(?:fail(?:ed|ure)?|error|timeout|timed\s+out)\b", summary):
+                return True
+            if re.search(r"\b(?:failed|failure|error|timeout|timed\s+out)\b", summary):
+                return True
+        return False
 
     @staticmethod
     def _is_external_agent_block(result: AgentResult) -> bool:

{engineering_platform-2.3.67 → engineering_platform-2.3.70}/src/engineering_platform/platform_version.py

@@ -12,7 +12,7 @@ from .storage import ENGINEERING_STORAGE_SCHEMA_VERSION
 
 SEMVER = re.compile(r"^(0|[1-9]\d*)\.(0|[1-9]\d*)\.(0|[1-9]\d*)$")
 CONTRACT = re.compile(r"^(\d{4})\.(0[1-9]|1[0-2])$")
-CURRENT_PLATFORM_VERSION = "2.3.67"
+CURRENT_PLATFORM_VERSION = "2.3.70"
 MANIFEST_FIELDS = frozenset(
     {
         "platform_version",

{engineering_platform-2.3.67 → engineering_platform-2.3.70}/src/engineering_platform/templates/workspace-config.json

@@ -3,7 +3,7 @@
   "platform": {
     "id": "engineering-platform",
     "name": "Engineering Platform",
-    "version": "2.3.67",
+    "version": "2.3.70",
     "generation": 2,
     "documentation_namespace": "engineering-platform",
     "capability_registry_version": 1

{engineering_platform-2.3.67 → engineering_platform-2.3.70}/src/engineering_platform.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: engineering-platform
-Version: 2.3.67
+Version: 2.3.70
 Summary: Local-first Engineering Platform execution operations runtime
 Author: Peter van Tol
 License: MIT