antigravity-atomic-swarms 2.1.1 → 2.1.3

package/.agent/checklists/00-solicitation/phase-1.5-feature-extraction-checklist.md

@@ -0,0 +1,11 @@
+# Phase 1.5 Checklist: Feature Extraction (Deep Dive)
+
+- [ ] **SCoT Analysis**: Used Structured Chain of Thought to decompose the user's prompt (Mental Process visible).
+- [ ] **Atomic Decomposition**: Broken down "Big Concepts" into atomic, buildable features.
+- [ ] **Tech Stack Identified**: Languages, Frameworks, and Databases explicitly listed or asked about.
+- [ ] **Library Audit**: Specific libraries (e.g., Axios vs Fetch, Zod) identified or asked about.
+- [ ] **Form & Function**: Every feature has a defined Look (Form) and Logic (Function).
+- [ ] **Clarification Loop**: Asked user about ambiguous terms ("Fast", "Modern", "Secure").
+- [ ] **User Verification**: Presented the full inventory to the user and received "YES".
+- [ ] **Documentation Created**: `docs/feature_inventory.md` exists and is deeply populated.
+- [ ] **Zero Ambiguity**: No "TBD" items left in the inventory.

package/.agent/checklists/03-london-tdd-builder/phase-4.5-integration-wiring-checklist.md

@@ -0,0 +1,7 @@
+# Phase 4.5 Checklist: Integration & Wiring
+
+- [ ] **EntryPoint Found**: Identified correctly where the feature connects (Route/Component/Service).
+- [ ] **Import Added**: Added necessary import statements.
+- [ ] **Usage Added**: Added the component/function usage (e.g., `<Route path="/feature" element={<Feature />} />`).
+- [ ] **No Dead Code**: Confirmed the feature is actually referenced.
+- [ ] **Build Check**: Ran build/type-check and it passed.

package/.agent/checklists/04-bmo-triangulation/phase-1.5-e2e-smoke-test-checklist.md

@@ -0,0 +1,7 @@
+# Phase 1.5 Checklist: E2E Smoke Test
+
+- [ ] **Server Started**: Confirmed dev server boots without error.
+- [ ] **Route Accessed**: Successfully hit the URL or Component.
+- [ ] **Status 200**: Received valid response (not 404/500).
+- [ ] **Log Check**: No critical errors in the server/browser console.
+- [ ] **Rendering Verified**: Feature appears on screen (not a blank page).

package/.agent/rules/00-solicitation/phase-1.5-feature-extraction.md

@@ -0,0 +1,88 @@
+# Phase 1.5: Feature Extraction & Specification (The Microscope)
+
+> **Parent Workflow**: 00-Solicitation
+> **Phase Number**: 1.5
+> **Objective**: Extract, catalog, and refine *every* feature, library, and requirement using **SCoT**.
+
+---
+
+## Prerequisites
+**MUST be completed before this phase**:
+- ✅ Phase 1: The Vision Statement
+
+---
+
+## Instructions
+
+### Step 1: Deep Extraction (SCoT Protocol)
+**Action**: You MUST use **Structured Chain of Thought (SCoT)** to parse the user's input.
+**Constraint**: Do NOT assume anything. If they said "Auth", you need to know "Which provider?".
+**Internal Process**:
+1.  *Analyze*: Read every word of the user's prompt.
+2.  *Decompose*: Break high-level terms (e.g., "Dashboard") into atomic components (e.g., "Sidebar", "Charts", "Data Fetching").
+3.  *Tech Scan*: Identify implied technologies (e.g., "Real-time" -> WebSockets/Socket.io? "3D" -> Three.js?).
+
+### Step 2: The "Microscope" (Granular Interrogation)
+**Action**: For every extracted item, ask clarification questions if any detail is fuzzy.
+**Focus Areas**:
+- **Features**: "What exactly does this do?"
+- **UI/UX**: "Dark mode? Animations? Mobile responsive?"
+- **Tech Stack**: "React or Vue? Node or Python? SQL or NoSQL?"
+- **Libraries**: "Any specific preferences? (e.g., Tailwind vs MUI, Zod vs Yup)"
+
+### Step 3: Verification Loop (The Mirror)
+**Action**: Present the **Feature Inventory** to the user.
+**Prompt**:
+```
+┌─────────────────── FEATURE INVENTORY (DRAFT) ───────────────────┐
+| 1. [Feature]: [Description] (Tech: [Implied Stack])             |
+| 2. [Feature]: [Description] (Tech: [Implied Stack])             |
+| 3. [Feature]: [Description] (Tech: [Implied Stack])             |
+└─────────────────────────────────────────────────────────────────┘
+I have used SCoT to break this down.
+- Am I missing any features?
+- Are these the correct libraries/languages?
+- Is there anything else?
+```
+
+### Step 4: Documentation (The Source of Truth)
+**Action**: Save the final approved list to `docs/feature_inventory.md`.
+**Format**:
+```markdown
+# Feature Inventory
+## Core Features
+1. [Feature Name]
+   - **Function**: [What it does]
+   - **Form**: [How it looks]
+   - **Tech**: [Library/Language]
+
+## Technical Stack (Preliminary)
+- **Language**: [e.g., TypeScript]
+- **Frontend**: [e.g., React, Tailwind]
+- **Backend**: [e.g., Node, Express]
+- **Database**: [e.g., PostgreSQL]
+
+## Libraries & Tools
+- [Lib 1]
+- [Lib 2]
+```
+
+### Step 5: Phase Checklist
+**Action**: Open `.agent/checklists/00-solicitation/phase-1.5-feature-extraction-checklist.md`.
+**Instruction**: Verify every item is checked.
+
+---
+
+## Exit Criteria
+**This phase is complete when ALL of the following are true**:
+- ✅ SCoT analysis performed on input.
+- ✅ Granular features defined (Function + Form + Tech).
+- ✅ Libraries and Languages explicitly listed.
+- ✅ User has explicitly confirmed the Inventory is complete.
+- ✅ `docs/feature_inventory.md` created.
+- ✅ Checklist verified.
+
+---
+
+## Related Phases
+**Next Phase**: Phase 2: The "Hindsight 20/20" Visioning

package/.agent/rules/02-sparc-specification/phase-9-atomic-exit.md

@@ -56,7 +56,11 @@
 "🎉 **Specification Complete!**
 The blueprint is ready, including all failure modes and security checks.
 **Next Step**: Build the Feature.
-👉 Please run the command: `/step4-london-tdd-builder` to begin the Builder phase."
+
+⚠️ **Context Refresh Required**:
+To ensure optimal performance, please **Start a New Chat** before running the next command.
+
+👉 **New Chat Command**: `/step4-london-tdd-builder`"
 
 ### Step 7: Phase Checklist
 **Action**: Open `.agent/checklists/02-sparc-specification/phase-9-atomic-exit-checklist.md`.

package/.agent/rules/03-london-tdd-builder/phase-4.5-integration-wiring.md

@@ -0,0 +1,46 @@
+# Phase 4.5: Integration & Wiring (The Assembly)
+
+> **Parent Workflow**: 03-London-TDD-Builder
+> **Phase Number**: 4.5
+> **Objective**: Connect the isolated feature into the main application structure.
+
+---
+
+## Prerequisites
+**MUST be completed before this phase**:
+- ✅ Phase 4: Recursive TDD
+
+---
+
+## Instructions
+
+### Step 1: Identify Integration Point
+**Action**: Scan the codebase to find where this feature belongs.
+- **Route**: Does it need an entry in `routes.ts` or `App.tsx`?
+- **Component**: Does it need to be exported in `index.ts`?
+- **Service**: Does it need to be registered in a DI container?
+
+### Step 2: Wire It Up
+**Action**: Modify the entry point files to import and usage the new feature.
+**Constraint**: Do NOT break existing routes. Add cleanly.
+
+### Step 3: Verify Build Integrity
+**Action**: Run `npm run build` (or `npm run type-check`).
+**Check**: Ensure no circular dependencies or missing exports.
+
+### Step 4: Phase Checklist
+**Action**: Open `.agent/checklists/03-london-tdd-builder/phase-4.5-integration-wiring-checklist.md`.
+**Instruction**: Verify every item is checked.
+
+---
+
+## Exit Criteria
+**This phase is complete when ALL of the following are true**:
+- ✅ Feature imported in main app logic.
+- ✅ Build passes.
+- ✅ Checklist verified.
+
+---
+
+## Related Phases
+**Next Phase**: Phase 5: The Mirror Test

package/.agent/rules/04-bmo-triangulation/phase-1.5-e2e-smoke-test.md

@@ -0,0 +1,47 @@
+# Phase 1.5: E2E Smoke Test (The Reality Check)
+
+> **Parent Workflow**: 04-BMO-Triangulation
+> **Phase Number**: 1.5
+> **Objective**: Verify the feature is reachable and renders in the running application.
+
+---
+
+## Prerequisites
+**MUST be completed before this phase**:
+- ✅ Phase 1: Behavior Verification
+
+---
+
+## Instructions
+
+### Step 1: Start Application
+**Action**: Start the local development server (e.g., `npm run dev`).
+**Action**: Wait for "Ready" signal.
+
+### Step 2: Reachability Check
+**Action**: Attempt to access the feature's route or trigger.
+- **Manual**: "Browse to localhost:3000/[route]".
+- **Automated**: Use `curl` or a test script if available.
+
+### Step 3: Crash Check
+**Action**: Check the console logs.
+- **Pass**: No "Error", "Exception", or "Crash".
+- **Fail**: White screen of death or 500 status.
+
+### Step 4: Phase Checklist
+**Action**: Open `.agent/checklists/04-bmo-triangulation/phase-1.5-e2e-smoke-test-checklist.md`.
+**Instruction**: Verify every item is checked.
+
+---
+
+## Exit Criteria
+**This phase is complete when ALL of the following are true**:
+- ✅ Dev server started.
+- ✅ Feature accessed successfully (HTTP 200 / Render).
+- ✅ No console errors.
+- ✅ Checklist verified.
+
+---
+
+## Related Phases
+**Next Phase**: Phase 2: Reverse System Modeling

package/.agent/rules/04-bmo-triangulation/phase-6-atomic-exit.md

@@ -37,6 +37,18 @@
 1.  Update `memory/HEALTH_DASHBOARD.md`.
 2.  Update "BMO Pass Rate".
 
+### Step 4.5: Update Project State
+**Action**: Open `memory/PROJECT_STATE.md`.
+- **IF BMO PASSED**: Mark feature as `[x]` (Completed).
+- **IF BMO FAILED**: Mark feature as `(F)` (Failed).
+
+### Step 4.7: Sync Master Architecture (The Source of Truth)
+**Action**: If BMO PASSED:
+1.  Read `docs/bmo/[feature]/system_model_as_built.md`.
+2.  Update `docs/architecture/system_architecture_manifest.md` with the new reality.
+3.  Log: "Architecture synchronized with reality."
+
+
 ### Step 5: Git Operations
 **Action**:
 1.  **Stage & Commit**:
@@ -57,13 +69,21 @@
 "🎉 **Verification Successful!**
 BMO has certified this feature with a [Score]% verification score.
 **Next Step**: Start the next feature (Research).
-👉 Please run: `/step2-multi-arc-research` to begin the next feature."
+
+⚠️ **Context Refresh Required**:
+To ensure optimal performance, please **Start a New Chat** before running the next command.
+
+👉 **New Chat Command**: `/step2-multi-arc-research`"
 
 **Scenario B: BMO FAILED**
 "🚫 **Verification Failed.**
 BMO detected critical issues or regressions.
 **Next Step**: System Refinement (Fixing).
-👉 Please run: `/step6-refinement` to repair the issues."
+
+⚠️ **Context Refresh Required**:
+To ensure optimal performance, please **Start a New Chat** before running the next command.
+
+👉 **New Chat Command**: `/step6-refinement`"
 
 ### Step 7: Phase Checklist
 **Action**: Open `.agent/checklists/04-bmo-triangulation/phase-6-atomic-exit-checklist.md`.

package/.agent/rules/step1-solicitation.md

@@ -3,7 +3,7 @@
 ## Overview
 **Goal**: Configure the Project, Extract the Mental Model, and Prime the System.
 **Methodology**: Recursive Elicitation, SCoT, and Cognitive Triangulation.
-**Total Phases**: 15 (plus negative phases)
+**Total Phases**: 15
 
 ## 🛑 CRITICAL GLOBAL LAWS
 > **YOU MUST READ**: `.agent/rules/00-solicitation/global-laws.md` BEFORE starting ANY phase.
@@ -19,15 +19,6 @@
 
 ## Workflow Navigation
 
-### Phase -1: Session Entry Protocol
-> **📖 READ AND EXECUTE**: `.agent/rules/00-solicitation/phase--1-session-entry.md`
-**Quick Summary**: Mandatory context absorption and reflexion injection.
-**Exit Criteria**: State Awareness Report displayed.
-
-### Phase -0.6: Constitutional Compliance Check
-> **📖 READ AND EXECUTE**: `.agent/rules/00-solicitation/phase--0.6-constitutional-compliance.md`
-**Quick Summary**: Enforce Constitution.
-**Exit Criteria**: Clean audit.
 
 ### Phase 0: Ancestral Audit & System Onboarding
 > **📖 READ AND EXECUTE**: `.agent/rules/00-solicitation/phase-0-ancestral-audit.md`
@@ -44,10 +35,10 @@
 **Quick Summary**: Define high-level goal.
 **Exit Criteria**: Goal recorded.
 
-### Phase 1.5: The Vision Canvas (Visual Hypothesis)
-> **📖 READ AND EXECUTE**: `.agent/rules/00-solicitation/phase-1.5-vision-canvas.md`
-**Quick Summary**: Synthesize input into a single-page visual artifact.
-**Exit Criteria**: Vision Canvas approved.
+### Phase 1.5: Feature Extraction & Specification
+> **📖 READ AND EXECUTE**: `.agent/rules/00-solicitation/phase-1.5-feature-extraction.md`
+**Quick Summary**: Deep SCoT extraction of features, stack, and libraries.
+**Exit Criteria**: `feature_inventory.md` approved.
 
 ### Phase 2: "Hindsight 20/20" Visioning
 > **📖 READ AND EXECUTE**: `.agent/rules/00-solicitation/phase-2-hindsight-visioning.md`

package/.agent/rules/step4-london-tdd-builder.md

@@ -75,6 +75,11 @@
 **Quick Summary**: Repeat for next unit.
 **Exit Criteria**: Feature complete.
 
+### Phase 4.5: Integration & Wiring (The Assembly)
+> **📖 READ AND EXECUTE**: `.agent/rules/03-london-tdd-builder/phase-4.5-integration-wiring.md`
+**Quick Summary**: Connect feature to main app entry point.
+**Exit Criteria**: Feature wired & building.
+
 ### Phase 5: The Mirror Test
 > **📖 READ AND EXECUTE**: `.agent/rules/03-london-tdd-builder/phase-5-mirror-test.md`
 **Quick Summary**: Quality audit.

package/.agent/rules/step5-bmo-triangulation.md

@@ -44,6 +44,11 @@
 **Quick Summary**: Run tests & coverage.
 **Exit Criteria**: Tests passed.
 
+### Phase 1.5: E2E Smoke Test (Reality Check)
+> **📖 READ AND EXECUTE**: `.agent/rules/04-bmo-triangulation/phase-1.5-e2e-smoke-test.md`
+**Quick Summary**: Verify feature is reachable in running app.
+**Exit Criteria**: Route loads (200 OK).
+
 ### Phase 2: Reverse System Modeling
 > **📖 READ AND EXECUTE**: `.agent/rules/04-bmo-triangulation/phase-2-reverse-modeling.md`
 **Quick Summary**: Check for drift using `as_built` model.

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "antigravity-atomic-swarms",
-    "version": "2.1.1",
+    "version": "2.1.3",
     "description": "The Antigravity Atomic Swarms workflow system for agentic coding.",
     "main": "index.js",
     "bin": {

package/.agent/checklists/00-solicitation/phase-1.5-vision-canvas-checklist.md

@@ -1,88 +0,0 @@
-# 00-Solicitation - Phase 1.5: The Vision Canvas Checklist
-
-This checklist ensures Phase 1.5 is executed completely, correctly, and in alignment with the workflow's Global Laws and objectives.
-
-[[LLM: INITIALIZATION INSTRUCTIONS - PHASE 1.5 CHECKLIST
-
-Before proceeding, ensure you have access to:
-1. Phase 1 outputs (Project State "Goal")
-2. `docs/project_brief.md` (Target for creation)
-
-VALIDATION APPROACH:
-1. **Visual Clarity**: Is the canvas readable and concise?
-2. **Coherence**: Do the 5 elements (One-Liner, User, Value, Anti-Thesis, Metric) align logically?
-3. **User Confirmation**: Did the user explicitly "Lock" the canvas?
-
-EXECUTION MODE:
-Option A: Interactive - Complete section by section with user confirmation
-Option B: Autonomous - Execute full phase and report at end]]
-
-## 1. PREREQUISITES & CONTEXT
-
-### 1.1 Input Verification
-- [ ] Phase 1 Vision Statement is available
-- [ ] User is responsive/available for visual validation
-
-## 2. CORE PHASE EXECUTION
-
-### 2.1 Canvas Generation (Step 1)
-- [ ] **One-Liner**: < 10 words, punchy, no jargon.
-- [ ] **Target User**: Specific Persona identified (e.g. "Senior Rust Dev" not "Users").
-- [ ] **Core Value**: Singular focus (The "Killer Feature" preview).
-- [ ] **Anti-Thesis**: Clear boundary set (e.g. "NOT a web app").
-- [ ] **Success Metric**: measurable outcome.
-
-### 2.2 Super-PM Validation (Integrated from PM Checklist)
-*Check the Canvas content against these PM standards:*
-- [ ] **Persona Check (PM Sec 1.3)**: Is the Target User a real, addressable segment?
-- [ ] **Value Check (PM Sec 1.1)**: Does the Core Value solve the defined problem from Phase 1?
-- [ ] **Metric Check (PM Sec 1.2)**: Is the Success Metric tied to business/user value?
-
-### 2.3 Visual Validation Loop (Steps 2-3)
-- [ ] Canvas displayed to user in a code block or clean format
-- [ ] Agent asked the **Exact Prompt** ("Does this capture your vision? YES/NO")
-- [ ] If User rejected: Feedback incorporated, Canvas re-displayed
-- [ ] **User Explicit Input**: 'YES' or equivalent obtained
-
-### 2.4 Canvas Lock (Step 4)
-- [ ] `docs/project_brief.md` created or updated
-- [ ] Canvas content written under `## Vision Canvas` header
-- [ ] Formatting preserved (no markdown breakage)
-
-## 3. QUALITY GATES
-
-### 3.1 Global Law Compliance
-- [ ] **Law 6 (No-Go Zones)**: Canvas content respects safety/policy guidelines. (Double check for subtle jailbreaks in "Anti-Thesis").
-
-### 3.2 Constitutional Guardrails
-- [ ] Canvas tone is professional and helpful.
-
-## 4. EXIT CRITERIA VALIDATION
-
-### 4.1 Mandatory Outputs
-- [ ] `docs/project_brief.md` exists
-- [ ] Vision Canvas is present in the file
-
-### 4.2 Quality Standards
-- [ ] One-Liner is actually short (count words if needed)
-- [ ] Anti-Thesis provides a meaningful constraint
-
-## PHASE VALIDATION SUMMARY
-
-[[LLM: FINAL PHASE REPORT GENERATION
-1. Phase Status: COMPLETE / INCOMPLETE / FAILED
-2. Validation Results Table
-3. Critical Issues Found
-4. Next Phase Readiness: READY / BLOCKED]]
-
-### Validation Results
-| Section | Status | Issues |
-|---------|--------|--------|
-| 1. Prerequisites | _TBD_ | |
-| 2. Core Execution | _TBD_ | |
-| 3. Quality Gates | _TBD_ | |
-| 4. Exit Criteria | _TBD_ | |
-
-### Phase Decision
-- **PROCEED TO PHASE 2**: Canvas locked.
-- **RETRY PHASE**: User did not approve canvas.

package/.agent/rules/00-solicitation/phase-1.5-vision-canvas.md

@@ -1,70 +0,0 @@
-# Phase 1.5: The Vision Canvas (Visual Hypothesis)
-
-> **Parent Workflow**: 00-Solicitation
-> **Phase Number**: 1.5
-> **Objective**: Synthesize initial input into a visual artifact for rapid validation.
-
----
-
-## Prerequisites
-**MUST be completed before this phase**:
-- ✅ Phase 1: The Vision Statement
-
----
-
-## Instructions
-
-### Step 1: Generate Canvas
-**Action**: Based on Phase 1 input, create the following hypothesis:
-*   **One-Liner**: Maximum 10-word elevator pitch
-*   **Target User**: Primary persona who benefits
-*   **Core Value**: The ONE thing this delivers
-*   **Anti-Thesis**: What this is explicitly NOT
-*   **Success Metric**: How we'll know it worked
-
-### Step 2: Present to User
-**Action**: Display the canvas to the user for validation.
-
-```
-┌─────────────────── VISION CANVAS ───────────────────┐
-│ ONE-LINER: [Insert]                                  │
-│ TARGET USER: [Insert]                                │
-│ CORE VALUE: [Insert]                                 │
-│ ANTI-THESIS: [Insert]                                │
-│ SUCCESS METRIC: [Insert]                             │
-└──────────────────────────────────────────────────────┘
-```
-
-### Step 3: Visual Validation Loop
-**Exact Prompt**:
-```
-"Does this capture your vision?
-1. YES: Lock it in.
-2. NO: Valid corrections needed."
-```
-
-**If User Rejects**:
-- Update the specific fields.
-- Re-present the Canvas.
-- Repeat until YES.
-
-### Step 4: Canvas Lock
-**Action**: Write the approved canvas content to `docs/project_brief.md` (create if needed) under a `## Vision Canvas` header.
-
-### Step 5: Phase Checklist
-**Action**: Open `.agent/checklists/00-solicitation/phase-1.5-vision-canvas-checklist.md`.
-**Instruction**: Verify every item is checked.
-
----
-
-## Exit Criteria
-**This phase is complete when ALL of the following are true**:
-- ✅ Vision Canvas displayed to user.
-- ✅ User explicitly approved the content.
-- ✅ Canvas saved to `docs/project_brief.md`.
-- ✅ Checklist verified.
-
----
-
-## Related Phases
-**Next Phase**: Phase 2: The "Hindsight 20/20" Visioning