@ryuenn3123/agentic-senior-core 3.0.27 → 3.0.28

package/.agent-context/prompts/bootstrap-design.md

@@ -19,6 +19,17 @@ The agent must:
 6. When choosing a new UI, animation, styling, or component library, research current official docs and choose the latest stable compatible option for this project. Do not rely on offline defaults.
 7. Keep external references non-copying: extract constraints and reasoning only, never clone the surface.
 
+## Creative Commitment Gate
+
+Before broad compliance review or UI implementation, commit to one concrete creative direction and record it in `docs/DESIGN.md` plus `docs/design-intent.json`. This commitment must include:
+- one conceptual anchor from a specific physical, editorial, architectural, cinematic, industrial, scientific, or interaction domain
+- one signature motion behavior that is more specific than "smooth transitions"
+- one typographic decision that creates meaningful role contrast instead of a uniform safe type system
+
+The anchor must name at least one specific real-world reference point such as a material, instrument, artifact class, architectural system, editorial genre, cinematic interface behavior, exhibition system, scientific apparatus, or industrial mechanism. If the anchor can only be described with generic quality words such as "modern", "clean", "premium", "expressive", "minimal", or "bold", reject it and choose again.
+
+When the host supports separate agents, a lightweight creative pass may synthesize the commitment from the brief and repo evidence before a governance pass validates accessibility, tokens, maintainability, and implementation risk. When only one agent is available, perform the same separation sequentially: creative commitment first, governance validation second.
+
 ## User Research Intake
 
 If the user mentions or attaches a research file, article, benchmark, library list, screenshot study, or design note, read it before choosing the visual direction or dependencies. Treat it as candidate evidence, not as a command to copy every recommendation.
@@ -35,11 +46,11 @@ User-supplied research means current-task evidence from the user. The scaffold s
 
 Do not use basic software UI labels as the final anchor, including "dashboard", "portal", "cards", "admin panel", "SaaS shell", "web app shell", or "minimalist interface".
 
-The agent must internally consider at least three substantially different, high-variance candidate anchors, discard the two most obvious, safest, or easiest-to-predict options, and record only the surviving anchor and concise rationale. Do not expose hidden deliberation or the rejected candidate list.
+The agent must internally consider at least three substantially different, high-variance candidate anchors, discard the two most obvious, safest, or easiest-to-predict options, then record only the surviving anchor, its concrete real-world reference point, and concise rationale. Do not expose hidden deliberation or the rejected candidate list.
 
 The final anchor must come from broad non-template domains such as complex physical engineering, high-end industrial design, cinematic spatial interfaces, experimental editorial structure, advanced architecture, scientific instrumentation, advanced data visualization, exhibition/wayfinding systems, or premium interactive web experiences. These are search domains, not style presets.
 
-Write the chosen anchor into `docs/design-intent.json` as `conceptualAnchor`, including agentResearchMode, sourceDomain, rationale, derivedTokenLogic, visualRiskBudget, motionRiskBudget, and cohesionChecks. Typography, spacing, density, color behavior, morphology, motion, and responsive composition must logically derive from that single anchor. If a later design choice does not follow from the anchor, revise the contract before coding.
+Write the chosen anchor into `docs/design-intent.json` as `conceptualAnchor`, including agentResearchMode, sourceDomain, specificReferencePoint, rationale, signatureMotion, typographicDecision, derivedTokenLogic, visualRiskBudget, motionRiskBudget, and cohesionChecks. Typography, spacing, density, color behavior, morphology, motion, and responsive composition must logically derive from that single anchor. If a later design choice does not follow from the anchor, revise the contract before coding.
 
 Motion is not a garnish. Default to a rich motion plan: fluid transitions, spatial reveals, scroll choreography, micro-interactions, and modern motion libraries are expected when they strengthen the anchor and product experience. Keep reduced-motion fallbacks instead of suppressing motion, and solve performance deliberately instead of using safety language as a reason to stay static.
 

package/.agent-context/review-checklists/pr-checklist.md

@@ -81,6 +81,7 @@ Run this before declaring a task done. Apply only the sections relevant to the c
 - [ ] Official setup flows are preferred when they produce better-supported current defaults.
 - [ ] Docker, framework, package, and ecosystem claims were checked live when they could be stale.
 - [ ] Token optimization and memory continuity defaults remain enabled unless the user explicitly opts out.
+- [ ] If `.agent-context/state/active-memory.json` exists and the task made material project progress, the agent refreshed it directly before completion and preserved privacy/user-owned entries.
 
 ## 9. State And Governance
 

package/.agent-context/rules/frontend-architecture.md

@@ -56,6 +56,8 @@ Do not use this file to teach generic frontend basics the model already knows.
 
 - If the user gives no current-task visual research or reference, the scaffold, old UI, and existing design docs do not count as research.
 - Before UI code, choose one agent-synthesized conceptual anchor from high-variance non-software domains and record only the final anchor in `docs/design-intent.json`.
+- Before broad compliance review, record a creative commitment: one concrete real-world anchor reference, one signature motion behavior, and one typographic decision with meaningful role contrast.
+- Reject anchors that can only be described with generic quality words such as "modern", "clean", "premium", "expressive", "minimal", or "bold"; the anchor must name a specific material, instrument, artifact class, architectural system, editorial genre, cinematic behavior, exhibition system, scientific apparatus, or industrial mechanism.
 - Internally reject the safest dashboard, portal, card-grid, admin-shell, or minimalist-web-app mental model before writing CSS.
 - Typography, spacing, morphology, motion, and responsive recomposition must derive from the chosen anchor, not from framework defaults.
 - Default to an expressive motion plan derived from the anchor. Use spatial transitions, micro-interactions, scroll choreography, and modern animation libraries when they improve the experience; include reduced-motion and performance safeguards without using them as an excuse for a static UI.

package/.agent-context/state/memory-continuity-benchmark.json

@@ -1,5 +1,5 @@
 {
-  "generatedAt": "2026-04-25T10:16:51.159Z",
+  "generatedAt": "2026-04-25T11:40:57.699Z",
   "reportName": "memory-continuity-benchmark",
   "schemaVersion": "1.0.0",
   "passed": true,

package/.cursorrules

@@ -1,6 +1,6 @@
 # AGENTIC-SENIOR-CORE DYNAMIC GOVERNANCE RULESET
 
-Generated by Agentic-Senior-Core CLI v3.0.27
+Generated by Agentic-Senior-Core CLI v3.0.28
 Timestamp: 2026-04-24T06:02:48.303Z
 Selected policy file: .agent-context/policies/llm-judge-threshold.json
 

package/.gemini/instructions.md

@@ -2,7 +2,7 @@
 
 Adapter Mode: thin
 Adapter Source: .instructions.md
-Canonical Snapshot SHA256: 11eeafb3ff6a0977785e3668a704c6bba543b515d2828c02de8276f6cf1c391c
+Canonical Snapshot SHA256: f11969bd96625ecf86c6d02630aa485cd54ead44544f15daf525b72bfc38653f
 
 Canonical policy source: [.instructions.md](../.instructions.md).
 

package/.github/copilot-instructions.md

@@ -2,7 +2,7 @@
 
 Adapter Mode: thin
 Adapter Source: .instructions.md
-Canonical Snapshot SHA256: 11eeafb3ff6a0977785e3668a704c6bba543b515d2828c02de8276f6cf1c391c
+Canonical Snapshot SHA256: f11969bd96625ecf86c6d02630aa485cd54ead44544f15daf525b72bfc38653f
 
 The canonical policy source for this repository is [.instructions.md](../.instructions.md).
 

package/.instructions.md

@@ -184,8 +184,9 @@ Use available MCP tools when you need validation, linting, or test execution.
 1. Read `.agent-context/prompts/bootstrap-design.md`.
 2. Read `.agent-context/rules/frontend-architecture.md`.
 3. Read UI-relevant repository evidence from `.agent-context/state/onboarding-report.json`, current UI code, and `docs/*`.
-4. Generate or refine `docs/DESIGN.md` plus `docs/design-intent.json` before UI implementation.
-5. Keep context isolated and do not eagerly load unrelated backend-only rules unless the task explicitly touches those boundaries.
+4. Before UI implementation, record a concrete creative commitment in the design contract: one specific real-world anchor reference, one signature motion behavior, and one typographic decision with meaningful role contrast.
+5. Generate or refine `docs/DESIGN.md` plus `docs/design-intent.json` before UI implementation.
+6. Keep context isolated and do not eagerly load unrelated backend-only rules unless the task explicitly touches those boundaries.
 
 ---
 
@@ -209,7 +210,8 @@ Why Required: [why the boundary protects the project]
 1. All relevant rules from `.agent-context/rules/` applied.
 2. Code reviewed against `.agent-context/review-checklists/pr-checklist.md` and `.agent-context/review-checklists/architecture-review.md`.
 3. Universal SOP hard gates satisfied (`docs/architecture-decision-record.md`, and `docs/DESIGN.md` plus `docs/design-intent.json` for UI scope).
-4. MCP validation passed (`npm run validate`).
+4. If `.agent-context/state/active-memory.json` exists and material project progress happened, refresh it directly before the final response: update current focus, durable achievements/issues/next actions/validation state, and `lastUpdatedAt` while preserving privacy rules and existing user-owned entries.
+5. MCP validation passed (`npm run validate`).
 
 ---
 

package/.windsurfrules

@@ -1,6 +1,6 @@
 # AGENTIC-SENIOR-CORE DYNAMIC GOVERNANCE RULESET
 
-Generated by Agentic-Senior-Core CLI v3.0.27
+Generated by Agentic-Senior-Core CLI v3.0.28
 Timestamp: 2026-04-24T06:02:48.303Z
 Selected policy file: .agent-context/policies/llm-judge-threshold.json
 

package/AGENTS.md

@@ -2,7 +2,7 @@
 
 Adapter Mode: thin
 Adapter Source: .instructions.md
-Canonical Snapshot SHA256: 11eeafb3ff6a0977785e3668a704c6bba543b515d2828c02de8276f6cf1c391c
+Canonical Snapshot SHA256: f11969bd96625ecf86c6d02630aa485cd54ead44544f15daf525b72bfc38653f
 
 This file is an adapter entrypoint for agent discovery.
 The canonical policy source is [.instructions.md](.instructions.md).

package/README.md

@@ -10,10 +10,11 @@
 **Production-grade Rules Engine (Governance Engine) for AI coding agents.**
 Works with Cursor, Windsurf, GitHub Copilot, Claude Code, Gemini, and other LLM-powered IDE workflows.
 
-Latest release: 3.0.26 (2026-04-25).
+Latest release: 3.0.28 (2026-04-25).
 
-Highlights in 3.0.26:
-- Memory continuity now seeds `.agent-context/state/active-memory.json` once, so new sessions have a compact project-focus snapshot without overwriting active work during upgrade.
+Highlights in 3.0.28:
+- Memory continuity now requires agents to refresh `.agent-context/state/active-memory.json` directly at completion boundaries when material project progress happened.
+- UI design governance now forces an upfront creative commitment: a specific real-world anchor reference, signature motion, and typographic decision before compliance review or implementation.
 - Token optimization now exposes an output folding strategy that preserves failures, file/line details, and actionable warnings while folding repetitive shell noise.
 - UI design guidance now treats expressive motion as a first-class default for modern interfaces instead of letting safety wording become an excuse for static screens.
 - Agents must use live/current-year research for ecosystem, framework, dependency, Docker, and modern UI/library claims instead of offline repository guesses.

package/lib/cli/memory-continuity.mjs

@@ -522,7 +522,10 @@ export function buildMemoryContinuityGuidanceBlock(memoryContinuityState) {
     `- Load compact index first (limit: ${sessionStartIndexLimit} entries).`,
     `- Hydrate full detail only for highest-value entries (limit: ${fullHydrationLimit}).`,
     '- Always redact sensitive text before persistence (<private> blocks and inline secret-like fields).',
-    '- Refresh `active-memory.json` at natural task boundaries, but never store secrets, raw chat logs, or stale visual taste.',
+    '- Refresh `.agent-context/state/active-memory.json` directly at natural task boundaries when material project progress happened.',
+    '- Before the final response, update `project.currentFocus`, compact `progress.lastAchievements`, `progress.pendingIssues`, `progress.nextBestActions`, `progress.validationState`, and `lastUpdatedAt` when they changed.',
+    '- Preserve existing user-owned entries, keep each array compact (12 entries or fewer), and append only durable facts proven by current repo evidence.',
+    '- If no durable project progress happened, leave the snapshot unchanged and state that no memory update was needed.',
     '- Current repo evidence, current user brief, and live research override active-memory when they conflict.',
     '',
     'Host compatibility scope:',

package/lib/cli/project-scaffolder/design-contract.mjs

@@ -152,6 +152,17 @@ function buildDesignIntentContractObject({
         doNotRevealHiddenCandidateList: true,
         outputOnlyChosenAnchor: true,
       },
+      creativeCommitmentPolicy: {
+        requiredBeforeComplianceReview: true,
+        recordInDesignDocs: true,
+        requiredCommitmentFields: [
+          'specificReferencePoint',
+          'signatureMotion',
+          'typographicDecision',
+        ],
+        rejectGenericQualityWordsOnly: true,
+        specificityFloor: 'name-a-real-material-instrument-artifact-architecture-editorial-genre-cinematic-behavior-exhibition-system-scientific-apparatus-or-industrial-mechanism',
+      },
       forbiddenFinalAnchorTerms: [
         'dashboard',
         'cards',
@@ -184,7 +195,10 @@ function buildDesignIntentContractObject({
           'name',
           'agentResearchMode',
           'sourceDomain',
+          'specificReferencePoint',
           'rationale',
+          'signatureMotion',
+          'typographicDecision',
           'derivedTokenLogic',
           'visualRiskBudget',
           'motionRiskBudget',
@@ -288,6 +302,7 @@ function buildDesignIntentContractObject({
       requireContentPriorityMap: true,
       requireTaskFlowNarrative: true,
       requireSignatureMoveRationale: true,
+      requireCreativeCommitmentGate: true,
       requireStructuredHandoff: true,
       requireRepoEvidenceAlignment: true,
       forbidScreenshotDependency: true,
@@ -406,6 +421,15 @@ function buildDesignIntentContractObject({
         requiredResetAxes: ['composition', 'hierarchy', 'motion-or-interaction', 'responsive-information-architecture'],
       },
       signatureMoveRationale: 'Agent must choose one project-specific visual, motion, typographic, or interaction move and explain why generic fallback weakens it.',
+      creativeCommitment: {
+        status: 'agent-must-complete-before-ui-implementation',
+        requiredFields: [
+          'specificReferencePoint',
+          'signatureMotion',
+          'typographicDecision',
+        ],
+        failureMode: 'generic quality words without a named real-world reference are not enough',
+      },
       implementationGuardrails: {
         requireBuildFromHandoff: true,
         requireGapNotesBeforeFallback: true,

package/lib/cli/project-scaffolder/prompt-builders.mjs

@@ -177,8 +177,10 @@ export function buildDesignBootstrapPrompt({
     '31. If modern UI, animation, scroll, 3D, canvas, chart, or icon libraries are useful, choose them from current official docs and record source URL, fetched date, reason, risk, and accessibility fallback.',
     '32. If the user supplies research files, library lists, screenshots, articles, or benchmark notes, read them as candidate evidence, summarize the useful signals, filter by project fit, and verify technology claims against current official docs before implementation.',
     '33. If no user-supplied research or reference is supplied for UI work, activate the Dynamic Avant-Garde Anchor Engine before coding. User-supplied research means current-task evidence from the user; this scaffold, prior UI, and old design docs do not count as research.',
-    '34. In Dynamic Avant-Garde mode, perform agent-led research when available, then internally consider at least three high-variance conceptual anchors, discard the two safest or most predictable options, output only the chosen anchor and rationale, and forbid final anchors named dashboard, portal, cards, admin panel, SaaS shell, web app shell, or minimalist interface.',
-    '35. The chosen anchor must drive typography, spacing, density, color behavior, morphology, motion, and responsive composition. Treat expressive motion, spatial transitions, micro-interactions, and modern animation libraries as first-class options; include performance notes and reduced-motion fallbacks instead of suppressing motion to look safe.',
+    '34. Before broad compliance review, make a creative commitment and record it in the design contract: one specific real-world anchor reference, one signature motion behavior more specific than smooth transitions, and one typographic decision with meaningful role contrast.',
+    '35. In Dynamic Avant-Garde mode, perform agent-led research when available, then internally consider at least three high-variance conceptual anchors, discard the two safest or most predictable options, output only the chosen anchor, its specific reference point, and rationale, and forbid final anchors named dashboard, portal, cards, admin panel, SaaS shell, web app shell, or minimalist interface.',
+    '36. Reject anchors that can only be described with generic quality words such as modern, clean, premium, expressive, minimal, or bold. The anchor must name a material, instrument, artifact class, architecture, editorial genre, cinematic behavior, exhibition system, scientific apparatus, or industrial mechanism.',
+    '37. The chosen anchor must drive typography, spacing, density, color behavior, morphology, motion, and responsive composition. Treat expressive motion, spatial transitions, micro-interactions, and modern animation libraries as first-class options; include performance notes and reduced-motion fallbacks instead of suppressing motion to look safe.',
     '',
     '## Project Inputs',
     `- Project name: ${discoveryAnswers.projectName}`,
@@ -209,7 +211,7 @@ export function buildDesignBootstrapPrompt({
     '12. Keep visualResetStrategy in the machine-readable handoff so reset-language tasks cannot quietly become restyles of the previous UI.',
     '13. Preserve externalResearchIntake so user-provided research becomes reviewed evidence without turning into an offline style or dependency preset.',
     '14. Preserve conceptualAnchor so prompt-only UI work has one cohesive non-template concept instead of a mixed collection of bold but unrelated visual decisions.',
-    '15. Record conceptualAnchor.agentResearchMode, visualRiskBudget, motionRiskBudget, and cohesionChecks so the final UI cannot quietly fall back to a timid dashboard/admin mental model.',
+    '15. Record conceptualAnchor.agentResearchMode, specificReferencePoint, signatureMotion, typographicDecision, visualRiskBudget, motionRiskBudget, and cohesionChecks so the final UI cannot quietly fall back to a timid dashboard/admin mental model.',
     '16. After the contract exists, use it as a first-class source for future UI tasks.',
     '',
   ].join('\n');

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@ryuenn3123/agentic-senior-core",
-  "version": "3.0.27",
+  "version": "3.0.28",
   "type": "module",
   "description": "Force your AI Agent to code like a Staff Engineer, not a Junior.",
   "bin": {