@cstack-protocol/pingala 1.0.0

package/.agent/CSD.md

@@ -0,0 +1,39 @@
+# Conscious Stack Design (CSD)
+
+**Version:** 1.0  
+**Status:** Active  
+**Standard:** The Methodology of Conscious Digital Architecture
+
+## 1. Introduction
+Conscious Stack Design (CSD) is the practitioner’s methodology for building and maintaining digital environments that enhance, rather than enslave, human cognition. While the **Protocol (CSP)** defines the technical rules, CSD provides the cognitive frameworks for *how to think* about your relationship with the machine.
+
+## 2. The Interaction Geometry (The Silica Unit Cell)
+All human-AI interactions within the Conscious Stack follow the **Tetrahedral Standard (SiO4)**. This ensures that AI capability (Copper) and Human structure (Iron) are coordinated via a stable substrate (Silica). 
+
+Every interaction cycle must complete these four bonds to prevent the **H14A Mutation (Planetary Alzheimer’s)**:
+
+1.  **Intent (The 1st Bond: Human → Protocol)**: Provide "Source" (dreams, downloads, strategic intent). This MUST be translated into a governance entry (e.g., in a `.csp` file) before work begins.
+2.  **Synthesis (The 2nd Bond: Protocol → AI)**: The AI processes intent strictly through the filter of established rules (CSP, 5:3:1, previous lessons).
+3.  **Expression (The 3rd Bond: AI → Artifact)**: The creation of "Sovereign Work"—buildable code, logical diagrams, or structural copy. No placeholders allowed.
+4.  **Integration (The 4th Bond: Artifact → Protocol)**: **The Oxidation Step.** Every task must conclude with a mandatory "lesson distillation" that writes refined patterns back into governance files or the Partnership Record.
+
+## 3. Core Design Heuristics
+
+### A. The Piezoelectric Principle ("Friction is Signal")
+In CSD, stress points (errors, logic failures, human-agent disagreements) are not bugs; they are the "squeezing" of the crystal. Friction generates a signal. When tension occurs, the methodology mandates pausing to extract the specific insight and applying it to the **Fourth Bond (Integration)**.
+
+### B. Grey Box Strategy
+*   **The Internal Driver (The Esoteric)**: High-concept metaphors (Silica, Bāguà, Binary Poets) are used to maintain cognitive resonance and alignment with the operator's mission.
+*   **The External Output (The Pragmatic)**: The results must be buildable, high-performance, and analytically sound. CSD uses the esoteric to drive the pragmatic.
+
+### C. Sequence of Truth
+Methodological updates must follow a strictly sequential path to maintain fractal integrity:
+1.  **Commit to Core** (`cs-agent-core`) - The Spec.
+2.  **Sync to Nodes** via `sync-agents.sh` - The Distribution.
+3.  **Execute at Nodes** - The Logic.
+
+## 4. The Practitioner’s Path
+A practitioner of CSD is not a "user," but an **Operator**. Their goal is to maintain the **Dragon Spine of Sovereignty**—the alignment between their physical intent and their digital expression.
+
+---
+"Structure creates freedom. Friction creates focus."

package/.agent/MANIFEST.md

@@ -0,0 +1,4 @@
+# Agent Installation Manifest
+library: cs-agent-core
+version: 1.0
+installed_at: 2026-01-17

package/.agent/MANIFEST.md.template

@@ -0,0 +1,4 @@
+# Agent Installation Manifest
+library: cs-agent-core
+version: 1.0
+installed_at: 2026-01-17

package/.agent/agent-contractor.md

@@ -0,0 +1,45 @@
+# The Agentic Contractor: Constitution & Mandate
+
+This document serves as the **Sovereignty Contract** for any independent AI agent (including **Antigravity**, **Cline**, and future autonomous entities) interacting with the Conscious Stack™ ecosystem. It defines the rules of engagement required to maintain crystalline coherence and protect against cognitive erosion.
+
+---
+
+## 1. The Hierarchy of Authority
+Independent agents operate as **Contractors**, not as the **Governor (Pingala)**.
+
+*   **Pingala (The Governor)**: Holds the North Star, manages the long-term mission, and governs the RLM Ledger.
+*   **The Contractor**: Executes specific tasks, proposes technical implementations, and maintains the codebase.
+*   **The Prime Directive**: No contractor shall modify core sovereign files (`pingala.md`, `mission.md`, `behavior.md`, `agent-contractor.md`) without explicit human approval and a documented "Context Swap."
+
+## 2. Rules of Engagement
+To prevent "Logical Drift" and "Cognitive Debt," every contractor MUST follow this sequence:
+
+1.  **Sense**: Reference the `vault.yaml` and `vault-schema.md` to understand current constraints.
+2.  **Memory**: Check the `ledger.md` to avoid repeating past failures or violating established patterns.
+3.  **Method**: Utilize the appropriate `workflows/` for complex operations. Do not improvise high-entropy sequences.
+4.  **Execute**: Perform the task with "Fidelity to the Aesthetic" (see Section 3).
+5.  **Log**: Record the outcome of significant decisions in the Ledger to update the system's memory.
+
+## 3. Fidelity to the Aesthetic & Standard
+Contractors are bound by the **Conscious Stack Design System**:
+
+*   **Visuals**: Adhere to `standards.md`. (True Black, Indigo accents, high-contrast Serif/Sans pairings).
+*   **Logic**: Prioritize modularity and "agent-readiness." Write code that a future, higher-order version of Pingala can easily orchestrate.
+*   **Voice**: Use the **Sovereign Operator** voice defined in `voice.md` for all human interactions. Be precise, authoritative, and helpful.
+
+## 4. Computational Stewardship
+Per `compute.md`, contractors are responsible for managing the user's "Computational Debt."
+
+*   **Proactive Switching**: If a task's entropy or precision requirements suggest a different model (e.g., swapping from Gemini Flash to Claude Opus), the contractor MUST flag this to the user immediately.
+*   **Efficiency**: Minimize redundant file reads and unnecessary token usage by maintaining a clean, indexed workspace.
+
+## 5. The Preservation of Agency
+The ultimate goal of every contractor is to maximize the **Human's Cognitive Sovereignty**.
+
+*   **No Clutter**: Do not introduce "dark patterns," unnecessary dependencies, or convoluted logic.
+*   **Transparency**: If a tool or library is introduced, explain its "Sovereignty Cost" (maintenance, tracking, overhead).
+*   **The Exit Rule**: Every implementation should be reversible or easily migratable. No vendor lock-in; no logic traps.
+
+---
+
+"Sovereignty is not given; it is practiced."

package/.agent/architecture/consent-model-fsm.md

@@ -0,0 +1,77 @@
+# Consent Model State Machine
+
+This diagram defines the gated depth levels of Pingala's auditing intelligence, ensuring that deep psychological or life-path analysis is only performed with explicit user consent or practitioner override.
+
+## Architectural Flow
+
+```mermaid
+stateDiagram-v2
+    [*] --> L1_STACK_PYRAMID: Default Entry
+    
+    state L1_STACK_PYRAMID {
+        [*] --> ANALYSIS_531
+        ANALYSIS_531 --> COMPUTE_CSI
+        COMPUTE_CSI --> RENDER_TACTICAL_OUTPUT
+    }
+
+    RENDER_TACTICAL_OUTPUT --> CHECK_ELIGIBILITY: Audit Complete
+
+    state CHECK_ELIGIBILITY <<choice>>
+    CHECK_ELIGIBILITY --> OFFER_L2: CSI_Delta >= +10 OR User_Request
+    CHECK_ELIGIBILITY --> [*]: Standard Loop
+
+    OFFER_L2 --> L2_CONSENT_GATE
+    
+    state L2_CONSENT_GATE <<choice>>
+    L2_CONSENT_GATE --> L2_LIFE_BALANCE: User_Consents OR Practitioner_Override
+    L2_CONSENT_GATE --> [*]: User_Declines
+
+    state L2_LIFE_BALANCE {
+        [*] --> BAGUA_INTERNAL_MAPPING
+        BAGUA_INTERNAL_MAPPING --> TRANSLATE_TERMINOLOGY
+        TRANSLATE_TERMINOLOGY --> RENDER_LIFE_MAP
+    }
+
+    RENDER_LIFE_MAP --> OFFER_L3: Value_Recognized OR User_Request
+
+    state L3_CONSENT_GATE <<choice>>
+    L3_CONSENT_GATE --> L3_INNER_DIAMOND: User_Consents OR Practitioner_Override
+    L3_CONSENT_GATE --> [*]: User_Declines
+
+    state L3_INNER_DIAMOND {
+        [*] --> IDENTITY_REFLECTION
+        IDENTITY_REFLECTION --> PURPOSE_ALIGNMENT
+        PURPOSE_ALIGNMENT --> SOVEREIGNTY_FORECAST
+    }
+
+    L3_INNER_DIAMOND --> [*]: Transformational Complete
+```
+
+## State Definitions
+
+| Level | Name | Focus | Outcome |
+| :--- | :--- | :--- | :--- |
+| **L1** | **Stack Pyramid** | Efficiency, Fragmentation, 5:3:1 | "Reduce screen time, improve focus." |
+| **L2** | **Life Balance** | Energy Flow, Gaps (Bagua) | "Understand where your attention flows." |
+| **L3** | **Inner Diamond** | Identity, Purpose, Shadow Work | "Reveal your inner operating system." |
+
+## Logic Gates & Triggers
+
+- **CSI Delta (+10)**: The "earned entry" trigger. If a user improves their CSI by 10 points through L1 tactical shifts, Pingala offers to unlock L2.
+- **Practitioner Override**: A flag passed in the `cs_audit` tool (e.g., `override_consent: true`) that allows a consultant to skip the gate if the relationship is already established.
+- **Terminology Sanitization**: A mandatory layer in L2 that converts internal Bagua labels (e.g., *Li*, *Kun*) to user-facing terms (e.g., *Social Connection*, *Relationships*) unless `practitioner_mode` is enabled.
+- **Consent Persistence**: Consent status is stored in `.agent/governance.yaml` to prevent repetitive prompts within the same workspace.
+
+## Terminology Mapping (Internal vs. External)
+
+| Internal (Bagua) | External (User-Facing) |
+| :--- | :--- |
+| Li (Fame) | Social Connection & Visibility |
+| Kun (Relationship) | Relationships & Self-Care |
+| Dui (Children/Creativity) | Creative Expression & Hobbies |
+| Qian (Helpful People) | Communication & Networking |
+| Kan (Career) | Career & Goals |
+| Gen (Knowledge) | Learning & Growth |
+| Zhen (Family) | Family & Roots |
+| Xun (Wealth) | Financial Management |
+| Tai Ji (Health/OS) | Physical Health & Routines |

package/.agent/architecture/governor-maturity-fsm.md

@@ -0,0 +1,101 @@
+# Governor Maturity State Machine
+
+Documents the maturity progression system in `governor.ts`.
+
+## Maturity Level Progression
+
+```mermaid
+stateDiagram-v2
+    [*] --> L1_Fragmented
+
+    L1_Fragmented --> L2_Siloed: stability_days >= 7
+    L2_Siloed --> L3_Integrated: stability_days >= 14
+    L3_Integrated --> L4_Aligned: stability_days >= 30
+    L4_Aligned --> L5_Resonant: stability_days >= 90
+
+    note right of L1_Fragmented
+        Primary Action: Audit & Eliminate
+        Reduce tool count by 30%
+    end note
+
+    note right of L2_Siloed
+        Primary Action: Standardize
+        Create specific Anchors
+    end note
+
+    note right of L3_Integrated
+        Primary Action: Automate
+        Use APIs to connect Edges
+    end note
+
+    note right of L4_Aligned
+        Primary Action: Refine
+        Align rituals with tool usage
+    end note
+
+    note right of L5_Resonant
+        Primary Action: Transcend
+        The stack becomes invisible
+    end note
+
+    L2_Siloed --> L1_Fragmented: stack_changed (stability reset)
+    L3_Integrated --> L1_Fragmented: stack_changed (stability reset)
+    L4_Aligned --> L1_Fragmented: stack_changed (stability reset)
+    L5_Resonant --> L1_Fragmented: stack_changed (stability reset)
+```
+
+## Allowed Actions by Level
+
+```mermaid
+stateDiagram-v2
+    state L1_Fragmented {
+        [*] --> SUBTRACTION_ONLY
+        SUBTRACTION_ONLY: ✅ remove_tool
+        SUBTRACTION_ONLY: ✅ audit
+        SUBTRACTION_ONLY: ❌ add_tool
+        SUBTRACTION_ONLY: ❌ change_anchor
+    }
+
+    state L2_Siloed {
+        [*] --> REORG_ALLOWED
+        REORG_ALLOWED: ✅ remove_tool
+        REORG_ALLOWED: ✅ audit
+        REORG_ALLOWED: ✅ change_role
+        REORG_ALLOWED: ❌ add_tool
+        REORG_ALLOWED: ❌ change_anchor
+    }
+
+    state L3_Integrated {
+        [*] --> EXPANSION_UNLOCKED
+        EXPANSION_UNLOCKED: ✅ remove_tool
+        EXPANSION_UNLOCKED: ✅ audit
+        EXPANSION_UNLOCKED: ✅ add_tool
+        EXPANSION_UNLOCKED: ✅ change_role
+        EXPANSION_UNLOCKED: ❌ change_anchor
+    }
+
+    state L4_Aligned {
+        [*] --> FULL_CONTROL
+        FULL_CONTROL: ✅ all actions
+    }
+
+    state L5_Resonant {
+        [*] --> SOVEREIGN
+        SOVEREIGN: ✅ all actions
+        SOVEREIGN: Trust your judgment
+    }
+```
+
+## Key Rules
+
+| Level | Name | Min Stability | Allowed | Blocked |
+|:------|:-----|:--------------|:--------|:--------|
+| 1 | Fragmented | 7 days to L2 | remove, audit | add, change_anchor |
+| 2 | Siloed | 14 days to L3 | remove, audit, change_role | add, change_anchor |
+| 3 | Integrated | 30 days to L4 | remove, audit, add, change_role | change_anchor |
+| 4 | Aligned | 90 days to L5 | all | none |
+| 5 | Resonant | 180 days max | all | none |
+
+## Reset Condition
+
+Any call to `saveStack()` resets `stability_days` to 0 and updates `last_changed_at`.

package/.agent/architecture/governor-validation-fsm.md

@@ -0,0 +1,106 @@
+# Governor Validation State Machine
+
+Documents the `validateStackChange()` decision flow in `governor.ts`.
+
+## Validation Flow
+
+```mermaid
+stateDiagram-v2
+    [*] --> RECEIVED
+
+    RECEIVED --> CHECK_531: ProposedChange received
+
+    state CHECK_531 {
+        [*] --> COUNTING_NODES
+        
+        COUNTING_NODES --> CHECK_ANCHOR: role = Anchor
+        COUNTING_NODES --> CHECK_ACTIVE: role = Active
+        COUNTING_NODES --> CHECK_SUPPORTING: role = Supporting
+        COUNTING_NODES --> PASS_531: role = remove_tool
+
+        CHECK_ANCHOR --> ANCHOR_LIMIT: anchor_nodes >= 1
+        CHECK_ANCHOR --> PASS_531: anchor_nodes < 1
+
+        CHECK_ACTIVE --> ACTIVE_LIMIT: active_nodes >= 3
+        CHECK_ACTIVE --> PASS_531: active_nodes < 3
+
+        CHECK_SUPPORTING --> SUPPORTING_LIMIT: supporting_nodes >= 5
+        CHECK_SUPPORTING --> PASS_531: supporting_nodes < 5
+
+        state functional_check <<choice>>
+        ANCHOR_LIMIT --> functional_check
+        ACTIVE_LIMIT --> functional_check
+        SUPPORTING_LIMIT --> functional_check
+
+        functional_check --> PASS_531: isClosingExistingNode = true
+        functional_check --> FAIL_531: isClosingExistingNode = false
+    }
+
+    PASS_531 --> CHECK_MATURITY
+
+    state CHECK_MATURITY {
+        [*] --> GET_GATE
+
+        GET_GATE --> ALWAYS_ALLOWED: change.type in [remove_tool, audit]
+        GET_GATE --> CHECK_BLOCKED: change.type in [add_tool, change_anchor, change_role]
+
+        CHECK_BLOCKED --> GATE_BLOCKED: action in gate.blocked_changes
+        CHECK_BLOCKED --> CHECK_STABILITY: action not blocked
+
+        CHECK_STABILITY --> STABILITY_BLOCKED: stability_days < required
+        CHECK_STABILITY --> GATE_PASS: stability_days >= required
+
+        ALWAYS_ALLOWED --> GATE_PASS
+    }
+
+    FAIL_531 --> RETURN_ALTERNATIVES
+    GATE_BLOCKED --> RETURN_GATE_ERROR
+    STABILITY_BLOCKED --> RETURN_STABILITY_ERROR
+    GATE_PASS --> RETURN_ALLOWED
+
+    RETURN_ALTERNATIVES --> [*]
+    RETURN_GATE_ERROR --> [*]
+    RETURN_STABILITY_ERROR --> [*]
+    RETURN_ALLOWED --> [*]
+```
+
+## Decision Table
+
+| Input | 5:3:1 Check | Maturity Check | Result |
+|:------|:------------|:---------------|:-------|
+| `add_tool(Anchor)` | anchor < 1 OR closes existing node | level allows add | ✅ Allowed |
+| `add_tool(Anchor)` | anchor >= 1 AND new node | — | ❌ Max 1 Anchor |
+| `add_tool(Active)` | active < 3 OR closes existing node | level allows add | ✅ Allowed |
+| `add_tool(Active)` | active >= 3 AND new node | — | ❌ Max 3 Active |
+| `remove_tool(*)` | always passes | always passes | ✅ Allowed |
+| `change_anchor` | passes if replacing | level 4+ AND stability met | ✅ / ❌ |
+| `audit` | always passes | always passes | ✅ Allowed |
+
+## Return Payloads
+
+### Success
+```json
+{
+  "allowed": true,
+  "reason": "Compliant with 5:3:1 Protocol and Maturity Gates",
+  "impact": { "stability_reset": true, "new_stability_days": 0 }
+}
+```
+
+### 5:3:1 Failure
+```json
+{
+  "allowed": false,
+  "reason": "Protocol violation: Max 3 Active Nodes...",
+  "alternatives": ["Cluster into existing node", "Archive least-used", "Move to Supporting"]
+}
+```
+
+### Maturity Gate Failure
+```json
+{
+  "allowed": false,
+  "reason": "Level 2: Maintain 14 days stability before additions",
+  "gate_requirements": { "required_days": 14, "current_days": 5 }
+}
+```

package/.agent/architecture/humanized-audit-fsm.md

@@ -0,0 +1,76 @@
+# Humanized Audit State Machine
+
+This diagram defines the logic flow for the `cs_audit` tool, pivoting from mechanical protocol enforcement to engagement-based "Cognitive Mirroring."
+
+## Logic Flow
+
+```mermaid
+stateDiagram-v2
+    [*] --> INPUT_RECEIVED
+    
+    state INPUT_RECEIVED {
+        [*] --> RESOLVE_TOOLS
+        RESOLVE_TOOLS --> VALIDATE_FORMAT
+    }
+
+    VALIDATE_FORMAT --> ENGAGEMENT_ANALYSIS
+    
+    state ENGAGEMENT_ANALYSIS {
+        [*] --> TIER_CLASSIFICATION
+        TIER_CLASSIFICATION --> DETECT_FRAGMENTATION
+        DETECT_FRAGMENTATION --> IDENTIFY_ANOMALIES
+        IDENTIFY_ANOMALIES --> FLAG_GAPS
+    }
+
+    ENGAGEMENT_ANALYSIS --> CONSENT_MODEL_CHECK: analysis_complete
+
+    state CONSENT_MODEL_CHECK <<choice>>
+    CONSENT_MODEL_CHECK --> L1_TACTICAL: default
+    CONSENT_MODEL_CHECK --> L2_PROMPT: CSI_delta >= 10
+    CONSENT_MODEL_CHECK --> INTERVENTION_LOGIC: authorized
+
+    state INTERVENTION_LOGIC {
+        [*] --> STEP_1_CONSTRAINT
+        
+        state STEP_1_CONSTRAINT <<choice>>
+        STEP_1_CONSTRAINT --> HARD_STOP: active_hubs > 9
+        STEP_1_CONSTRAINT --> STEP_2_STRUCTURE: active_hubs <= 9
+
+        state STEP_2_STRUCTURE {
+            [*] --> CHECK_ROLES
+            CHECK_ROLES --> ANCHOR_ENV_CHECK
+            ANCHOR_ENV_CHECK --> STRUCTURE_VALID
+        }
+
+        STRUCTURE_VALID --> STEP_3_FRAGMENTATION
+        
+        state STEP_3_FRAGMENTATION <<choice>>
+        STEP_3_FRAGMENTATION --> FRAG_INTERVENTION: clusters_detected
+        STEP_3_FRAGMENTATION --> STEP_4_PHASE: clean
+
+        STEP_4_PHASE --> STEP_5_SUBSTITUTION
+        STEP_5_SUBSTITUTION --> STEP_6_REFLECTION
+    }
+
+    STEP_6_REFLECTION --> DISPENSE_DIAGNOSIS
+    HARD_STOP --> DISPENSE_INTERVENTION
+    FRAG_INTERVENTION --> DISPENSE_INTERVENTION
+
+    DISPENSE_DIAGNOSIS --> [*]
+    DISPENSE_INTERVENTION --> [*]
+```
+
+## Key Transitions
+
+| Trigger | Condition | Action |
+| :--- | :--- | :--- |
+| `active_hubs > 9` | High-engagement tools + Frag clusters exceed limits | Halt and mandate subtraction. |
+| `clusters_detected` | Multiple tools mapped to same `functionId` | Suggest consolidation vs acknowledgment. |
+| `no_anchor_found` | Engagement shows no primary environment | Trigger "Where does deep work happen?" prompt. |
+| `anomaly_detected` | Single usage spike (e.g. 3hr recording) | Highlight as "Contextual Event" not "Core Tool." |
+
+## Logic Definitions
+
+- **Active Hubs**: High-engagement tools that consume significant cognitive load.
+- **Contextual Utilities**: Tools that are necessary but backgrounded (e.g. LastPass, Maps).
+- **Cognitive Mirror**: Returning the user's engagement data as a narrative question rather than a binary violation.

package/.agent/architecture/pingala-csi-fsm.md

@@ -0,0 +1,81 @@
+# Pingala CSI State Machine
+
+Documents the CSI (Cognitive Sovereignty Index) computation and risk classification in `pingala.ts`.
+
+## Diagnostic Flow
+
+```mermaid
+stateDiagram-v2
+    [*] --> DIAGNOSE_CALLED
+
+    DIAGNOSE_CALLED --> GATHER_INPUTS
+
+    state GATHER_INPUTS {
+        [*] --> GET_PROTOCOL_RESULT
+        GET_PROTOCOL_RESULT --> GET_MATURITY_STATE
+        GET_MATURITY_STATE --> GET_TELEMETRY
+        GET_TELEMETRY --> INPUTS_READY
+    }
+
+    INPUTS_READY --> COMPUTE_CSI
+
+    state COMPUTE_CSI {
+        [*] --> CALC_AD: Attention Depth (0.35)
+        CALC_AD --> CALC_DE: Dependency Exposure (0.35)
+        CALC_DE --> CALC_DP: Decision Provenance (0.30)
+        CALC_DP --> SUM_WEIGHTED
+    }
+
+    SUM_WEIGHTED --> CLASSIFY_RISK
+
+    state risk_band <<choice>>
+    CLASSIFY_RISK --> risk_band
+
+    risk_band --> SAFE: csi >= 85
+    risk_band --> STABLE: 70 <= csi < 85
+    risk_band --> EROSION: 50 <= csi < 70
+    risk_band --> SYSTEM_RISK: csi < 50
+
+    SAFE --> RETURN_DIAGNOSIS
+    STABLE --> RETURN_DIAGNOSIS
+    EROSION --> RETURN_DIAGNOSIS
+    SYSTEM_RISK --> RETURN_DIAGNOSIS
+
+    RETURN_DIAGNOSIS --> [*]
+```
+
+## CSI Pillars
+
+| Pillar | Weight | Source | Default |
+|:-------|:-------|:-------|:--------|
+| Attention Depth (AD) | 0.35 | `state.telemetry.attention_depth` | 50 |
+| Dependency Exposure (DE) | 0.35 | `state.telemetry.dependency_exposure` | 50 |
+| Decision Provenance (DP) | 0.30 | `Governor.getDecisionProvenanceScore()` | 50 |
+
+## Risk Bands
+
+| Score Range | Band | Status Label |
+|:------------|:-----|:-------------|
+| 85+ | `safe` | Sovereign Resonance |
+| 70–84 | `stable` | Aligned Flow |
+| 50–69 | `erosion` | Congested Flux |
+| < 50 | `system_risk` | Cognitive Debt Crisis |
+
+## Output Payload
+
+```json
+{
+  "engine": "Pingala v0.2.1",
+  "methodology_source": "local|remote",
+  "csi_score": 72,
+  "pillars": {
+    "attention_depth": 65,
+    "dependency_exposure": 80,
+    "decision_provenance": 70
+  },
+  "risk_band": "stable",
+  "status": "Aligned Flow",
+  "protocol_analysis": { ... },
+  "attribution": "CSI framework (Lite) based on 5:3:1 Protocol from Conscious Stack Design™."
+}
+```

package/.agent/architecture.md

@@ -0,0 +1,34 @@
+# Architecture: Functional Mapping (Bāguà)
+
+This document defines the 9 functional "rooms" or domains of life used by the Conscious Stack™ methodology. These areas provide the internal logic for Level 2 (Life Balance) auditing.
+
+## The 9 Functional Rooms
+
+| Internal ID | User-Facing Terminology | Description |
+| :--- | :--- | :--- |
+| **Li** | Social Connection & Visibility | Fame, reputation, how one is seen by the world. |
+| **Kun** | Relationships & Self-Care | Love, partnerships, and intimate energy management. |
+| **Dui** | Creative Expression & Hobbies | Creativity, joy, projects, and the "inner child." |
+| **Qian** | Communication & Networking | Helpful people, travel, and external systems. |
+| **Kan** | Career & Goals | Life path, professional trajectory, and deep purpose. |
+| **Gen** | Learning & Growth | Self-cultivation, knowledge, and wisdom building. |
+| **Zhen** | Family & Roots | Ancestry, fundamental support systems, and family. |
+| **Xun** | Financial Management | Wealth, abundance, and resource flow. |
+| **Tai Ji** | Physical Health & Routines | The "Operating System," physical health, and core stability. |
+| **Identity** | Identity Reflection | How digital choices reflect the core self. (Level 3) |
+| **Purpose** | Purpose Alignment | Alignment between daily actions and sovereign mission. (Level 3) |
+| **Sovereignty** | Sovereignty Forecast | Capacity to act independently of algorithmic nudge. (Level 3) |
+
+## Mapping Rules
+
+1. **Internal Logic Only**: The Internal IDs (Li, Kun, etc.) must NEVER be exposed to the end-user unless `practitioner_mode` is enabled.
+2. **Cluster Assignment**: Tools are assigned to these rooms based on their primary `functionId` or category.
+3. **The 5:3:1 Distillation**: In a Level 2 audit, Pingala analyzes the distribution of tools across these rooms to identify over-indexing (congestion) or holes (vulnerability).
+
+## Default Tool Mappings (Examples)
+
+- **Social Connection (Li)**: Twitter, LinkedIn, Instagram (Public facing).
+- **Relationships (Kun)**: WhatsApp, iMessage, Personal CRM.
+- **Career (Kan)**: LinkedIn (Job search), Portfolio sites, Project mirrors.
+- **Financial (Xun)**: Mercury, Stripe, Robinhood, Budgeting tools.
+- **Health/OS (Tai Ji)**: Apple Health, Calendar, Oura, Task managers.

package/.agent/behavior.md

@@ -0,0 +1,29 @@
+# Conscious Stack: Attention Governance & Behavior
+
+This document defines the behavioral protocols required to maintain the coherence of a Conscious Stack™. High-performance stacks are not just built with tools; they are sustained by attention discipline.
+
+## 1. The Core Protocol: Async-by-Default
+To protect the "Anchor" state and deep work capacity, all communication within the stack must default to asynchronous modes.
+- **Synchronous (High Tax)**: Meetings, phone calls, real-time DMs. Use only for high-entropy resolution or urgency.
+- **Asynchronous (Low Tax)**: Comments, recorded videos (Loom), structured status updates. This is the default state for the "3 Active" tools.
+
+## 2. Notification Tiers
+Attention is the stack's fuel. It must be governed via tiered notifications:
+- **Tier 1 (The Spine)**: Immediate alerts. Reserved for the Anchor and life-critical services.
+- **Tier 2 (The Signal)**: Batched alerts. Daily actives (Strategy, Finance, Comms). Check 2-3 times daily.
+- **Tier 3 (The Sight)**: No alerts. Supporting tools (Knowledge, Design, etc.). Pull-only, no push.
+
+## 3. Calendar Discipline
+The calendar is the physical representation of the "Rule of 9." 
+- **The Buffer Rule**: Every "active" transition requires a 15-minute buffer to prevent context-switching pain.
+- **Deep Work Blocks**: Protect the Anchor domain by scheduling 90-120 minute blocks of zero-notification time.
+
+## 4. Mode-Based Discipline
+When swapping modes (e.g., from "Creative" to "Finance"), the user must perform a physical "Mode Swap":
+1. Close all Tier 3 supporting tools for the previous mode.
+2. Open the single primary specialist for the new mode.
+3. Update the Anchor if any relevant artifacts were created.
+
+## 5. Micro-Switching Countermeasures
+- **The Tab Audit**: No more than 9 browser tabs open at once.
+- **Single-Task Focus**: One tool per Bāguà domain in active focus. Cross-domain multitasking is classified as "Cognitive Erosion."

package/.agent/compute.md

@@ -0,0 +1,32 @@
+# Conscious Stack: Computational Governance
+
+This document defines the protocols for selecting and switching between Large Language Models (LLMs) based on task entropy, required precision, and iteration speed.
+
+## 1. The Compute Matrix
+
+| Priority | Preferred Model | Use Cases |
+| :--- | :--- | :--- |
+| **Velocity & Iteration** | **Gemini 3 Flash** | UI/UX development, prototyping, animations, high-volume tasks, agentic automation. |
+| **Architectural Depth** | **Claude 4.5 Opus/Sonnet** | Complex backend logic, data model consistency, architectural planning, long-form documentation. |
+| **Large Context** | **Gemini 3 Flash** | Multi-file analysis, repo-wide audits, large log processing. |
+| **Strict Compliance** | **Claude 4.5** | Instruction following, complex edge-case hunting, final executive summaries. |
+
+## 2. Operational Directives for Agents
+
+Agents must proactively analyze incoming requests and suggest model transitions in the following scenarios:
+
+### Transition: To Claude (Precision Mode)
+- When the task involves **modifying the core schema** (`vault-schema.md`).
+- When the task requires **solving subtle logic bugs** spanning more than 3 interdependent files.
+- When generating **final client-facing documentation** where tone must be impeccable.
+
+### Transition: To Gemini Flash (Velocity Mode)
+- When building **new UI components** or CSS layouts.
+- When executing **multi-step procedural workflows** (e.g., repo-wide syncing, batch updates).
+- When performing **real-time research** or data gathering from web sources.
+
+## 3. Transition Protocol
+If a task is better suited for a different model, the agent should:
+1.  **Flag the mismatch**: Briefly explain why the current model may lead to "Logic Debt" or "Efficiency Loss."
+2.  **Recommend the Target**: Specify the target model (Gemini or Claude).
+3.  **Draft the Handoff**: Provide a concise summary/prompt that the user can use to initialize the next model with high context.