Improvement: Consolidated email provider creation, validation, and connection testing behind one canonical Email Manager path shared by runtime sending, test email, and admin connection checks.
Fix: Corrected free/pro notification gating so premium providers and PRO notification types are enforced consistently in UI, save sanitization, and runtime delivery.
Fix: Repaired email queue persistence and queue operations so new emails append correctly, sent/failed history is retained, and manual process/retry/clear actions operate on the real queue state.
Improvement: Refactored Tools and maintenance operations behind one shared Tools service so the Tools page, log viewer, and maintenance AJAX handlers use the same summary and execution path.
Fix: Corrected service-container loading for newly registered services by adding explicit class-file loading during service registration and health checks.
Fix: Replaced the old global redirect-notice queue with page-scoped admin flash notices so action messages appear only on the page that triggered them instead of leaking across Dashboard, Tools, Notifications, and other plugin screens.
1.8.12
Improvement: Unified Products step advanced filters under one canonical conditions contract shared by UI, validation, and runtime execution.
Improvement: Simplified Products step browser-side validation and moved contradiction and business-rule enforcement onto the server for more consistent results.
Fix: Corrected advanced-filter operator handling, condition execution, and price/shipping semantics so filtered product targeting behaves reliably.
Improvement: Refactored Campaign Health and dashboard health aggregation into dedicated diagnostics services with cleaner, deterministic inputs.
Improvement: Added one shared campaign applicability validator so save, review, health, and Campaign Intelligence all use the same discount validity rules.
Fix: Corrected scope-aware validation, discounted-product counts, blocker messaging, storefront fixed-savings badges, and category fallback resolution for invalid campaign scopes.
Maintenance: Removed redundant review and health normalization paths after consolidating applicability and diagnostics flows.
1.8.11
Improvement: Replaced the old additive Campaign Intelligence confidence heuristic with an evidence-weighted model based on signal completeness, data freshness, and cross-signal consistency.
Improvement: Added confidence reasons to the Intelligence reliability payload so confidence bands are explainable across dashboard, overview, and wizard surfaces.
Improvement: Added builder-side recommendation quality guards so weak generic primary actions are replaced before output in low-evidence contexts.
Improvement: Added evidence-backed action why explanations and wired them through the shared Intelligence payload, view model, and dashboard next-step block.
Improvement: Expanded Campaign Intelligence decision telemetry across dashboard, overview, wizard, planner, and campaigns list contexts with fallback and why coverage tracking.
Improvement: Added an internal telemetry review summary to the Intelligence advanced/governance payload so weak contexts and underperforming reason buckets are easier to inspect.
Improvement: Updated dashboard Upcoming Opportunities routing so linked campaign actions use Intelligence deep-linking and telemetry-aware campaign URLs instead of raw edit links when recommendation metadata is available.
Improvement: Consolidated plugin debugging onto a single logging backend with one canonical plugin.log file for PHP, browser, AJAX, and Tools-page log viewing flows.
Improvement: Added centralized log configuration resolution so effective debug mode, log level, retention, and mirroring behavior are shared consistently across PHP services and localized JavaScript.
Improvement: Promoted log_manager to a shared container-managed service so the Tools page and log viewer use the same log-management path and configuration source.
Fix: Tightened low-reliability confidence behavior for stale, estimated, sparse, or contradictory campaign contexts to reduce overconfident recommendations.
Fix: Repaired browser-to-server debug logging by aligning AJAX action names, nonce handling, and logger routing so wizard and admin-side browser logs are written reliably.
Fix: Updated the Tools page log viewer/reporting flow to reflect effective logging state correctly, explain historical lower-severity entries, and keep plain-text log downloads unaltered during output sanitization.
Improvement: Consolidated persistent plugin caching under one cache-manager source of truth with normalized key namespaces for campaigns, products, analytics, reference data, settings, and dashboard data.
Improvement: Integrated WooCommerce badge payload and badge HTML caching into the shared cache manager so badge rendering follows the same versioning and invalidation flow as the rest of the plugin cache.
Improvement: Hardened automatic cache invalidation for campaign, product, settings, license, dashboard, and currency-change events with safer persistent-object-cache fallback behavior.
Fix: Corrected grouped object-cache invalidation and eliminated stale mixed cache paths that could leave outdated campaign or badge data behind on sites using persistent object cache.
Maintenance: Removed redundant cache layers, duplicate invalidation hooks, broken ad hoc cache-manager construction paths, and unused cache helpers after the cache-system consolidation.
Maintenance: Removed the final legacy PHP logging facade along with redundant debug logger, console logger, duplicate AJAX handlers, and unused log-viewer branches after consolidating the logging architecture.
Fix: Tightened free/pro packaging so free builds exclude Pro implementation files while keeping upgrade/promo UI surfaces.
Fix: Updated WordPress.org readme short-description structure to avoid importer warning and improve listing consistency.
Maintenance: Release packaging and metadata alignment for 1.8.1.
1.8.0
Major release: Campaign Intelligence 1.8.0 introduces a full operational decision layer for WooCommerce promotions. This version shifts the dashboard from scattered metrics to action-oriented guidance so teams can quickly answer: “What should I do right now?” with clearer risk signals, next-step recommendations, reliability context, and disciplined execution boundaries.
New: Introduced a unified Campaign Intelligence engine with normalized contract outputs (risk_now, next_step, performance, advanced) and shared view-model shaping across dashboard, overview, and wizard surfaces.
New: Added dashboard-first Intelligence UX with 3 default blocks (Risk Now, Next Step, Performance), plus expandable details for evidence, diagnostics, and recommendation context.
New: Added first-class Intelligence metrics for override_intensity_percent, campaign_revenue_contribution_percent, and promotion_stability_score (structural health).
New: Added profile-driven Promotion Guardrails (Conservative, Balanced, Aggressive) with optional custom boundaries for max discount load %, max override intensity %, max concurrent campaigns, and minimum stability score.
New: Added soft-enforcement confirmation gates when actions exceed guardrail boundaries (safety confirmation, not hard blocking).
New: Added guardrail evaluation across launch and activation paths, including wizard launch, schedule-driven activation, dashboard/manual activation, and bulk activation.
New: Added override telemetry with temporary stability penalties after confirmed overrides, including dashboard visibility of active penalty state.
New: Added daily Intelligence snapshots with trust diffs (“Since yesterday…”) for trend-aware risk/performance context.
New: Added a compact Discipline status and 30-day Discipline Report with overrides in window, most breached boundary, average stability, and days within bounds.
Improvement: Updated terminology clarity (Decision Strength, Evidence Strength) and reduced default exposure of governance/internal wording in user-facing views.
Improvement: Added recommendation lifecycle visibility in Next Step states (still open, in progress, resolved).
Improvement: Added block-level detail expanders (Risk details, Why this recommendation, Performance details) to keep default UI focused and actionable.
Fix: Resolved risk-panel detail truncation by scoping bullet-limiting styles to summary bullets only.
Maintenance: Removed and reduced legacy Intelligence fields and UI paths no longer used by the 3-block decision contract.
