Merge pull request #9 from serversdwn/dev

Update to 0.6.0. Docs updated.

.gitignore

@@ -5,6 +5,7 @@ __pycache__/
 *.pyc
 *.log
 /.vscode/
+.vscode/
 # =============================
 # 🔐 Environment files (NEVER commit secrets!)
 # =============================

.vscode/settings.json

@@ -1,7 +0,0 @@
-{
-    "workbench.colorCustomizations": {
-        "activityBar.background": "#16340C",
-        "titleBar.activeBackground": "#1F4911",
-        "titleBar.activeForeground": "#F6FDF4"
-    }
-}

CHANGELOG.md

@@ -9,6 +9,105 @@ Format based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/) and [Se
 
 ---
 
+## [0.6.0] - 2025-12-18
+
+### Added - Autonomy System (Phase 1 & 2)
+
+**Autonomy Phase 1** - Self-Awareness & Planning Foundation
+- **Executive Planning Module** [cortex/autonomy/executive/planner.py](cortex/autonomy/executive/planner.py)
+  - Autonomous goal setting and task planning capabilities
+  - Multi-step reasoning for complex objectives
+  - Integration with self-state tracking
+- **Self-State Management** [cortex/data/self_state.json](cortex/data/self_state.json)
+  - Persistent state tracking across sessions
+  - Memory of past actions and outcomes
+  - Self-awareness metadata storage
+- **Self Analyzer** [cortex/autonomy/self/analyzer.py](cortex/autonomy/self/analyzer.py)
+  - Analyzes own performance and decision patterns
+  - Identifies areas for improvement
+  - Tracks cognitive patterns over time
+- **Test Suite** [cortex/tests/test_autonomy_phase1.py](cortex/tests/test_autonomy_phase1.py)
+  - Unit tests for phase 1 autonomy features
+
+**Autonomy Phase 2** - Decision Making & Proactive Behavior
+- **Autonomous Actions Module** [cortex/autonomy/actions/autonomous_actions.py](cortex/autonomy/actions/autonomous_actions.py)
+  - Self-initiated action execution
+  - Context-aware decision implementation
+  - Action logging and tracking
+- **Pattern Learning System** [cortex/autonomy/learning/pattern_learner.py](cortex/autonomy/learning/pattern_learner.py)
+  - Learns from interaction patterns
+  - Identifies recurring user needs
+  - Adapts behavior based on learned patterns
+- **Proactive Monitor** [cortex/autonomy/proactive/monitor.py](cortex/autonomy/proactive/monitor.py)
+  - Monitors system state for intervention opportunities
+  - Detects patterns requiring proactive response
+  - Background monitoring capabilities
+- **Decision Engine** [cortex/autonomy/tools/decision_engine.py](cortex/autonomy/tools/decision_engine.py)
+  - Autonomous decision-making framework
+  - Weighs options and selects optimal actions
+  - Integrates with orchestrator for coordinated decisions
+- **Orchestrator** [cortex/autonomy/tools/orchestrator.py](cortex/autonomy/tools/orchestrator.py)
+  - Coordinates multiple autonomy subsystems
+  - Manages tool selection and execution
+  - Handles NeoMem integration (with disable capability)
+- **Test Suite** [cortex/tests/test_autonomy_phase2.py](cortex/tests/test_autonomy_phase2.py)
+  - Unit tests for phase 2 autonomy features
+
+**Autonomy Phase 2.5** - Pipeline Refinement
+- Tightened integration between autonomy modules and reasoning pipeline
+- Enhanced self-state persistence and tracking
+- Improved orchestrator reliability
+- NeoMem integration refinements in vector store handling [neomem/neomem/vector_stores/qdrant.py](neomem/neomem/vector_stores/qdrant.py)
+
+### Added - Documentation
+
+- **Complete AI Agent Breakdown** [docs/PROJECT_LYRA_COMPLETE_BREAKDOWN.md](docs/PROJECT_LYRA_COMPLETE_BREAKDOWN.md)
+  - Comprehensive system architecture documentation
+  - Detailed component descriptions
+  - Data flow diagrams
+  - Integration points and API specifications
+
+### Changed - Core Integration
+
+- **Router Updates** [cortex/router.py](cortex/router.py)
+  - Integrated autonomy subsystems into main routing logic
+  - Added endpoints for autonomous decision-making
+  - Enhanced state management across requests
+- **Reasoning Pipeline** [cortex/reasoning/reasoning.py](cortex/reasoning/reasoning.py)
+  - Integrated autonomy-aware reasoning
+  - Self-state consideration in reasoning process
+- **Persona Layer** [cortex/persona/speak.py](cortex/persona/speak.py)
+  - Autonomy-aware response generation
+  - Self-state reflection in personality expression
+- **Context Handling** [cortex/context.py](cortex/context.py)
+  - NeoMem disable capability for flexible deployment
+
+### Changed - Development Environment
+
+- Updated [.gitignore](.gitignore) for better workspace management
+- Cleaned up VSCode settings
+- Removed [.vscode/settings.json](.vscode/settings.json) from repository
+
+### Technical Improvements
+
+- Modular autonomy architecture with clear separation of concerns
+- Test-driven development for new autonomy features
+- Enhanced state persistence across system restarts
+- Flexible NeoMem integration with enable/disable controls
+
+### Architecture - Autonomy System Design
+
+The autonomy system operates in layers:
+1. **Executive Layer** - High-level planning and goal setting
+2. **Decision Layer** - Evaluates options and makes choices
+3. **Action Layer** - Executes autonomous decisions
+4. **Learning Layer** - Adapts behavior based on patterns
+5. **Monitoring Layer** - Proactive awareness of system state
+
+All layers coordinate through the orchestrator and maintain state in `self_state.json`.
+
+---
+
 ## [0.5.2] - 2025-12-12
 
 ### Fixed - LLM Router & Async HTTP

README.md

@@ -1,10 +1,12 @@
-# Project Lyra - README v0.5.1
+# Project Lyra - README v0.6.0
 
-Lyra is a modular persistent AI companion system with advanced reasoning capabilities.
+Lyra is a modular persistent AI companion system with advanced reasoning capabilities and autonomous decision-making.
-It provides memory-backed chat using **NeoMem** + **Relay** + **Cortex**,
+It provides memory-backed chat using **Relay** + **Cortex** with integrated **Autonomy System**,
-with multi-stage reasoning pipeline powered by HTTP-based LLM backends.
+featuring a multi-stage reasoning pipeline powered by HTTP-based LLM backends.
 
-**Current Version:** v0.5.1 (2025-12-11)
+**Current Version:** v0.6.0 (2025-12-18)
+
+> **Note:** As of v0.6.0, NeoMem is **disabled by default** while we work out integration hiccups in the pipeline. The autonomy system is being refined independently before full memory integration.
 
 ## Mission Statement
 
@@ -24,7 +26,8 @@ Project Lyra operates as a **single docker-compose deployment** with multiple Do
 - OpenAI-compatible endpoint: `POST /v1/chat/completions`
 - Internal endpoint: `POST /chat`
 - Routes messages through Cortex reasoning pipeline
-- Manages async calls to NeoMem and Cortex ingest
+- Manages async calls to Cortex ingest
+- *(NeoMem integration currently disabled in v0.6.0)*
 
 **2. UI** (Static HTML)
 - Browser-based chat interface with cyberpunk theme
@@ -32,18 +35,20 @@ Project Lyra operates as a **single docker-compose deployment** with multiple Do
 - Saves and loads sessions
 - OpenAI-compatible message format
 
-**3. NeoMem** (Python/FastAPI) - Port 7077
+**3. NeoMem** (Python/FastAPI) - Port 7077 - **DISABLED IN v0.6.0**
 - Long-term memory database (fork of Mem0 OSS)
 - Vector storage (PostgreSQL + pgvector) + Graph storage (Neo4j)
 - RESTful API: `/memories`, `/search`
 - Semantic memory updates and retrieval
 - No external SDK dependencies - fully local
+- **Status:** Currently disabled while pipeline integration is refined
 
 ### Reasoning Layer
 
 **4. Cortex** (Python/FastAPI) - Port 7081
-- Primary reasoning engine with multi-stage pipeline
+- Primary reasoning engine with multi-stage pipeline and autonomy system
 - **Includes embedded Intake module** (no separate service as of v0.5.1)
+- **Integrated Autonomy System** (NEW in v0.6.0) - See Autonomy System section below
 - **4-Stage Processing:**
   1. **Reflection** - Generates meta-awareness notes about conversation
   2. **Reasoning** - Creates initial draft answer using context
@@ -82,9 +87,49 @@ Project Lyra operates as a **single docker-compose deployment** with multiple Do
 
 Each module can be configured to use a different backend via environment variables.
 
+### Autonomy System (NEW in v0.6.0)
+
+**Cortex Autonomy Subsystems** - Multi-layered autonomous decision-making and learning
+- **Executive Layer** [cortex/autonomy/executive/](cortex/autonomy/executive/)
+  - High-level planning and goal setting
+  - Multi-step reasoning for complex objectives
+  - Strategic decision making
+- **Decision Engine** [cortex/autonomy/tools/decision_engine.py](cortex/autonomy/tools/decision_engine.py)
+  - Autonomous decision-making framework
+  - Option evaluation and selection
+  - Coordinated decision orchestration
+- **Autonomous Actions** [cortex/autonomy/actions/](cortex/autonomy/actions/)
+  - Self-initiated action execution
+  - Context-aware behavior implementation
+  - Action logging and tracking
+- **Pattern Learning** [cortex/autonomy/learning/](cortex/autonomy/learning/)
+  - Learns from interaction patterns
+  - Identifies recurring user needs
+  - Adaptive behavior refinement
+- **Proactive Monitoring** [cortex/autonomy/proactive/](cortex/autonomy/proactive/)
+  - System state monitoring
+  - Intervention opportunity detection
+  - Background awareness capabilities
+- **Self-Analysis** [cortex/autonomy/self/](cortex/autonomy/self/)
+  - Performance tracking and analysis
+  - Cognitive pattern identification
+  - Self-state persistence in [cortex/data/self_state.json](cortex/data/self_state.json)
+- **Orchestrator** [cortex/autonomy/tools/orchestrator.py](cortex/autonomy/tools/orchestrator.py)
+  - Coordinates all autonomy subsystems
+  - Manages tool selection and execution
+  - Handles external integrations (with enable/disable controls)
+
+**Autonomy Architecture:**
+The autonomy system operates in coordinated layers, all maintaining state in `self_state.json`:
+1. Executive Layer → Planning and goals
+2. Decision Layer → Evaluation and choices
+3. Action Layer → Execution
+4. Learning Layer → Pattern adaptation
+5. Monitoring Layer → Proactive awareness
+
 ---
 
-## Data Flow Architecture (v0.5.1)
+## Data Flow Architecture (v0.6.0)
 
 ### Normal Message Flow:
 
@@ -97,11 +142,13 @@ Cortex (7081)
   ↓ (internal Python call)
 Intake module → summarize_context()
   ↓
+Autonomy System → Decision evaluation & pattern learning
+  ↓
 Cortex processes (4 stages):
   1. reflection.py → meta-awareness notes (CLOUD backend)
-  2. reasoning.py → draft answer (PRIMARY backend)
+  2. reasoning.py → draft answer (PRIMARY backend, autonomy-aware)
   3. refine.py → refined answer (PRIMARY backend)
-  4. persona/speak.py → Lyra personality (CLOUD backend)
+  4. persona/speak.py → Lyra personality (CLOUD backend, autonomy-aware)
   ↓
 Returns persona answer to Relay
   ↓
@@ -109,9 +156,11 @@ Relay → POST /ingest (async)
   ↓
 Cortex → add_exchange_internal() → SESSIONS buffer
   ↓
-Relay → NeoMem /memories (async, planned)
+Autonomy System → Update self_state.json (pattern tracking)
   ↓
 Relay → UI (returns final response)
+
+Note: NeoMem integration disabled in v0.6.0
 ```
 
 ### Cortex 4-Stage Reasoning Pipeline:
@@ -239,13 +288,13 @@ rag/
 All services run in a single docker-compose stack with the following containers:
 
 **Active Services:**
-- **neomem-postgres** - PostgreSQL with pgvector extension (port 5432)
-- **neomem-neo4j** - Neo4j graph database (ports 7474, 7687)
-- **neomem-api** - NeoMem memory service (port 7077)
 - **relay** - Main orchestrator (port 7078)
-- **cortex** - Reasoning engine with embedded Intake (port 7081)
+- **cortex** - Reasoning engine with embedded Intake and Autonomy System (port 7081)
 
-**Disabled Services:**
+**Disabled Services (v0.6.0):**
+- **neomem-postgres** - PostgreSQL with pgvector extension (port 5432) - *disabled while refining pipeline*
+- **neomem-neo4j** - Neo4j graph database (ports 7474, 7687) - *disabled while refining pipeline*
+- **neomem-api** - NeoMem memory service (port 7077) - *disabled while refining pipeline*
 - **intake** - No longer needed (embedded in Cortex as of v0.5.1)
 - **rag** - Beta Lyrae RAG service (port 7090) - currently disabled
 
@@ -278,7 +327,32 @@ The following LLM backends are accessed via HTTP (not part of docker-compose):
 
 ## Version History
 
-### v0.5.1 (2025-12-11) - Current Release
+### v0.6.0 (2025-12-18) - Current Release
+**Major Feature: Autonomy System (Phase 1, 2, and 2.5)**
+- ✅ Added autonomous decision-making framework
+- ✅ Implemented executive planning and goal-setting layer
+- ✅ Added pattern learning system for adaptive behavior
+- ✅ Implemented proactive monitoring capabilities
+- ✅ Created self-analysis and performance tracking system
+- ✅ Integrated self-state persistence (`cortex/data/self_state.json`)
+- ✅ Built decision engine with orchestrator coordination
+- ✅ Added autonomous action execution framework
+- ✅ Integrated autonomy into reasoning and persona layers
+- ✅ Created comprehensive test suites for autonomy features
+- ✅ Added complete system breakdown documentation
+
+**Architecture Changes:**
+- Autonomy system integrated into Cortex reasoning pipeline
+- Multi-layered autonomous decision-making architecture
+- Self-state tracking across sessions
+- NeoMem disabled by default while refining pipeline integration
+- Enhanced orchestrator with flexible service controls
+
+**Documentation:**
+- Added [PROJECT_LYRA_COMPLETE_BREAKDOWN.md](docs/PROJECT_LYRA_COMPLETE_BREAKDOWN.md)
+- Updated changelog with comprehensive autonomy system details
+
+### v0.5.1 (2025-12-11)
 **Critical Intake Integration Fixes:**
 - ✅ Fixed `bg_summarize()` NameError preventing SESSIONS persistence
 - ✅ Fixed `/ingest` endpoint unreachable code
@@ -320,17 +394,19 @@ The following LLM backends are accessed via HTTP (not part of docker-compose):
 
 ---
 
-## Known Issues (v0.5.1)
+## Known Issues (v0.6.0)
 
-### Critical (Fixed in v0.5.1)
+### Temporarily Disabled (v0.6.0)
-- ~~Intake SESSIONS not persisting~~ ✅ **FIXED**
+- **NeoMem disabled by default** - Being refined independently before full integration
-- ~~`bg_summarize()` NameError~~ ✅ **FIXED**
+  - PostgreSQL + pgvector storage inactive
-- ~~`/ingest` endpoint unreachable code~~ ✅ **FIXED**
+  - Neo4j graph database inactive
+  - Memory persistence endpoints not active
+- RAG service (Beta Lyrae) currently disabled in docker-compose.yml
 
 ### Non-Critical
 - Session management endpoints not fully implemented in Relay
-- RAG service currently disabled in docker-compose.yml
+- Full autonomy system integration still being refined
-- NeoMem integration in Relay not yet active (planned for v0.5.2)
+- Memory retrieval integration pending NeoMem re-enablement
 
 ### Operational Notes
 - **Single-worker constraint**: Cortex must run with single Uvicorn worker to maintain SESSIONS state
@@ -338,12 +414,14 @@ The following LLM backends are accessed via HTTP (not part of docker-compose):
 - Diagnostic endpoints (`/debug/sessions`, `/debug/summary`) available for troubleshooting
 
 ### Future Enhancements
+- Re-enable NeoMem integration after pipeline refinement
+- Full autonomy system maturation and optimization
 - Re-enable RAG service integration
 - Implement full session persistence
 - Migrate SESSIONS to Redis for multi-worker support
 - Add request correlation IDs for tracing
 - Comprehensive health checks across all services
-- NeoMem integration in Relay
+- Enhanced pattern learning with long-term memory integration
 
 ---
 
@@ -576,12 +654,16 @@ NeoMem is a derivative work based on Mem0 OSS (Apache 2.0).
 
 ## Development Notes
 
-### Cortex Architecture (v0.5.1)
+### Cortex Architecture (v0.6.0)
 - Cortex contains embedded Intake module at `cortex/intake/`
 - Intake is imported as: `from intake.intake import add_exchange_internal, SESSIONS`
 - SESSIONS is a module-level global dictionary (singleton pattern)
 - Single-worker constraint required to maintain SESSIONS state
 - Diagnostic endpoints available for debugging: `/debug/sessions`, `/debug/summary`
+- **NEW:** Autonomy system integrated at `cortex/autonomy/`
+  - Executive, decision, action, learning, and monitoring layers
+  - Self-state persistence in `cortex/data/self_state.json`
+  - Coordinated via orchestrator with flexible service controls
 
 ### Adding New LLM Backends
 1. Add backend URL to `.env`:

cortex/autonomy/Assembly-spec.md

@@ -0,0 +1,249 @@
+# 📐 Project Lyra — Cognitive Assembly Spec
+**Version:** 0.6.1  
+**Status:** Canonical reference  
+**Purpose:** Define clear separation of Self, Thought, Reasoning, and Speech
+
+---
+
+## 1. High-Level Overview
+
+Lyra is composed of **four distinct cognitive layers**, plus I/O.
+
+Each layer has:
+- a **responsibility**
+- a **scope**
+- clear **inputs / outputs**
+- explicit **authority boundaries**
+
+No layer is allowed to “do everything.”
+
+---
+
+## 2. Layer Definitions
+
+### 2.1 Autonomy / Self (NON-LLM)
+
+**What it is**
+- Persistent identity
+- Long-term state
+- Mood, preferences, values
+- Continuity across time
+
+**What it is NOT**
+- Not a reasoning engine
+- Not a planner
+- Not a speaker
+- Not creative
+
+**Implementation**
+- Data + light logic
+- JSON / Python objects
+- No LLM calls
+
+**Lives at**
+```
+project-lyra/autonomy/self/
+```
+
+**Inputs**
+- Events (user message received, response sent)
+- Time / idle ticks (later)
+
+**Outputs**
+- Self state snapshot
+- Flags / preferences (e.g. verbosity, tone bias)
+
+---
+
+### 2.2 Inner Monologue (LLM, PRIVATE)
+
+**What it is**
+- Internal language-based thought
+- Reflection
+- Intent formation
+- “What do I think about this?”
+
+**What it is NOT**
+- Not final reasoning
+- Not execution
+- Not user-facing
+
+**Model**
+- MythoMax
+
+**Lives at**
+```
+project-lyra/autonomy/monologue/
+```
+
+**Inputs**
+- User message
+- Self state snapshot
+- Recent context summary
+
+**Outputs**
+- Intent
+- Tone guidance
+- Depth guidance
+- “Consult executive?” flag
+
+**Example Output**
+```json
+{
+  "intent": "technical_exploration",
+  "tone": "focused",
+  "depth": "deep",
+  "consult_executive": true
+}
+```
+
+---
+
+### 2.3 Cortex (Reasoning & Execution)
+
+**What it is**
+- Thinking pipeline
+- Planning
+- Tool selection
+- Task execution
+- Draft generation
+
+**What it is NOT**
+- Not identity
+- Not personality
+- Not persistent self
+
+**Models**
+- DeepSeek-R1 → Executive / Planner
+- GPT-4o-mini → Executor / Drafter
+
+**Lives at**
+```
+project-lyra/cortex/
+```
+
+**Inputs**
+- User message
+- Inner Monologue output
+- Memory / RAG / tools
+
+**Outputs**
+- Draft response (content only)
+- Metadata (sources, confidence, etc.)
+
+---
+
+### 2.4 Persona / Speech (LLM, USER-FACING)
+
+**What it is**
+- Voice
+- Style
+- Expression
+- Social behavior
+
+**What it is NOT**
+- Not planning
+- Not deep reasoning
+- Not decision-making
+
+**Model**
+- MythoMax
+
+**Lives at**
+```
+project-lyra/core/persona/
+```
+
+**Inputs**
+- Draft response (from Cortex)
+- Tone + intent (from Inner Monologue)
+- Persona configuration
+
+**Outputs**
+- Final user-visible text
+
+---
+
+## 3. Message Flow (Authoritative)
+
+### 3.1 Standard Message Path
+
+```
+User
+  ↓
+UI
+  ↓
+Relay
+  ↓
+Cortex
+  ↓
+Autonomy / Self (state snapshot)
+  ↓
+Inner Monologue (MythoMax)
+  ↓
+[ consult_executive? ]
+    ├─ Yes → DeepSeek-R1 (plan)
+    └─ No  → skip
+  ↓
+GPT-4o-mini (execute & draft)
+  ↓
+Persona (MythoMax)
+  ↓
+Relay
+  ↓
+UI
+  ↓
+User
+```
+
+### 3.2 Fast Path (No Thinking)
+
+```
+User → UI → Relay → Persona → Relay → UI
+```
+
+---
+
+## 4. Authority Rules (Non-Negotiable)
+
+- Self never calls an LLM
+- Inner Monologue never speaks to the user
+- Cortex never applies personality
+- Persona never reasons or plans
+- DeepSeek never writes final answers
+- MythoMax never plans execution
+
+---
+
+## 5. Folder Mapping
+
+```
+project-lyra/
+├── autonomy/
+│   ├── self/
+│   ├── monologue/
+│   └── executive/
+├── cortex/
+├── core/
+│   └── persona/
+├── relay/
+└── ui/
+```
+
+---
+
+## 6. Current Status
+
+- UI ✔
+- Relay ✔
+- Cortex ✔
+- Persona ✔
+- Autonomy ✔
+- Inner Monologue ⚠ partially wired
+- Executive gating ⚠ planned
+
+---
+
+## 7. Next Decision
+
+Decide whether **Inner Monologue runs every message** or **only when triggered**.

cortex/autonomy/__init__.py

@@ -0,0 +1 @@
+# Autonomy module for Lyra

cortex/autonomy/actions/__init__.py

@@ -0,0 +1 @@
+"""Autonomous action execution system."""

cortex/autonomy/actions/autonomous_actions.py

@@ -0,0 +1,480 @@
+"""
+Autonomous Action Manager - executes safe, self-initiated actions.
+"""
+
+import logging
+import json
+from typing import Dict, List, Any, Optional
+from datetime import datetime
+
+logger = logging.getLogger(__name__)
+
+
+class AutonomousActionManager:
+    """
+    Manages safe autonomous actions that Lyra can take without explicit user prompting.
+
+    Whitelist of allowed actions:
+    - create_memory: Store information in NeoMem
+    - update_goal: Modify goal status
+    - schedule_reminder: Create future reminder
+    - summarize_session: Generate conversation summary
+    - learn_topic: Add topic to learning queue
+    - update_focus: Change current focus area
+    """
+
+    def __init__(self):
+        """Initialize action manager with whitelisted actions."""
+        self.allowed_actions = {
+            "create_memory": self._create_memory,
+            "update_goal": self._update_goal,
+            "schedule_reminder": self._schedule_reminder,
+            "summarize_session": self._summarize_session,
+            "learn_topic": self._learn_topic,
+            "update_focus": self._update_focus
+        }
+
+        self.action_log = []  # Track all actions for audit
+
+    async def execute_action(
+        self,
+        action_type: str,
+        parameters: Dict[str, Any],
+        context: Dict[str, Any]
+    ) -> Dict[str, Any]:
+        """
+        Execute a single autonomous action.
+
+        Args:
+            action_type: Type of action (must be in whitelist)
+            parameters: Action-specific parameters
+            context: Current context state
+
+        Returns:
+            {
+                "success": bool,
+                "action": action_type,
+                "result": action_result,
+                "timestamp": ISO timestamp,
+                "error": optional error message
+            }
+        """
+        # Safety check: action must be whitelisted
+        if action_type not in self.allowed_actions:
+            logger.error(f"[ACTIONS] Attempted to execute non-whitelisted action: {action_type}")
+            return {
+                "success": False,
+                "action": action_type,
+                "error": f"Action '{action_type}' not in whitelist",
+                "timestamp": datetime.utcnow().isoformat()
+            }
+
+        try:
+            logger.info(f"[ACTIONS] Executing autonomous action: {action_type}")
+
+            # Execute the action
+            action_func = self.allowed_actions[action_type]
+            result = await action_func(parameters, context)
+
+            # Log successful action
+            action_record = {
+                "success": True,
+                "action": action_type,
+                "result": result,
+                "timestamp": datetime.utcnow().isoformat(),
+                "parameters": parameters
+            }
+
+            self.action_log.append(action_record)
+            logger.info(f"[ACTIONS] Action {action_type} completed successfully")
+
+            return action_record
+
+        except Exception as e:
+            logger.error(f"[ACTIONS] Action {action_type} failed: {e}")
+
+            error_record = {
+                "success": False,
+                "action": action_type,
+                "error": str(e),
+                "timestamp": datetime.utcnow().isoformat(),
+                "parameters": parameters
+            }
+
+            self.action_log.append(error_record)
+            return error_record
+
+    async def execute_batch(
+        self,
+        actions: List[Dict[str, Any]],
+        context: Dict[str, Any]
+    ) -> List[Dict[str, Any]]:
+        """
+        Execute multiple actions sequentially.
+
+        Args:
+            actions: List of {"action": str, "parameters": dict}
+            context: Current context state
+
+        Returns:
+            List of action results
+        """
+        results = []
+
+        for action_spec in actions:
+            action_type = action_spec.get("action")
+            parameters = action_spec.get("parameters", {})
+
+            result = await self.execute_action(action_type, parameters, context)
+            results.append(result)
+
+            # Stop on first failure if critical
+            if not result["success"] and action_spec.get("critical", False):
+                logger.warning(f"[ACTIONS] Critical action {action_type} failed, stopping batch")
+                break
+
+        return results
+
+    # ========================================
+    # Whitelisted Action Implementations
+    # ========================================
+
+    async def _create_memory(
+        self,
+        parameters: Dict[str, Any],
+        context: Dict[str, Any]
+    ) -> Dict[str, Any]:
+        """
+        Create a memory entry in NeoMem.
+
+        Parameters:
+            - text: Memory content (required)
+            - tags: Optional tags for memory
+            - importance: 0.0-1.0 importance score
+        """
+        text = parameters.get("text")
+        if not text:
+            raise ValueError("Memory text required")
+
+        tags = parameters.get("tags", [])
+        importance = parameters.get("importance", 0.5)
+        session_id = context.get("session_id", "autonomous")
+
+        # Import NeoMem client
+        try:
+            from memory.neomem_client import store_memory
+
+            result = await store_memory(
+                text=text,
+                session_id=session_id,
+                tags=tags,
+                importance=importance
+            )
+
+            return {
+                "memory_id": result.get("id"),
+                "text": text[:50] + "..." if len(text) > 50 else text
+            }
+
+        except ImportError:
+            logger.warning("[ACTIONS] NeoMem client not available, simulating memory storage")
+            return {
+                "memory_id": "simulated",
+                "text": text[:50] + "..." if len(text) > 50 else text,
+                "note": "NeoMem not available, memory not persisted"
+            }
+
+    async def _update_goal(
+        self,
+        parameters: Dict[str, Any],
+        context: Dict[str, Any]
+    ) -> Dict[str, Any]:
+        """
+        Update goal status in self-state.
+
+        Parameters:
+            - goal_id: Goal identifier (required)
+            - status: New status (pending/in_progress/completed)
+            - progress: Optional progress note
+        """
+        goal_id = parameters.get("goal_id")
+        if not goal_id:
+            raise ValueError("goal_id required")
+
+        status = parameters.get("status", "in_progress")
+        progress = parameters.get("progress")
+
+        # Import self-state manager
+        from autonomy.self.state import get_self_state_instance
+
+        state = get_self_state_instance()
+        active_goals = state._state.get("active_goals", [])
+
+        # Find and update goal
+        updated = False
+        for goal in active_goals:
+            if isinstance(goal, dict) and goal.get("id") == goal_id:
+                goal["status"] = status
+                if progress:
+                    goal["progress"] = progress
+                goal["updated_at"] = datetime.utcnow().isoformat()
+                updated = True
+                break
+
+        if updated:
+            state._save_state()
+            return {
+                "goal_id": goal_id,
+                "status": status,
+                "updated": True
+            }
+        else:
+            return {
+                "goal_id": goal_id,
+                "updated": False,
+                "note": "Goal not found"
+            }
+
+    async def _schedule_reminder(
+        self,
+        parameters: Dict[str, Any],
+        context: Dict[str, Any]
+    ) -> Dict[str, Any]:
+        """
+        Schedule a future reminder.
+
+        Parameters:
+            - message: Reminder text (required)
+            - delay_minutes: Minutes until reminder
+            - priority: 0.0-1.0 priority score
+        """
+        message = parameters.get("message")
+        if not message:
+            raise ValueError("Reminder message required")
+
+        delay_minutes = parameters.get("delay_minutes", 60)
+        priority = parameters.get("priority", 0.5)
+
+        # For now, store in self-state's learning queue
+        # In future: integrate with scheduler/cron system
+        from autonomy.self.state import get_self_state_instance
+
+        state = get_self_state_instance()
+
+        reminder = {
+            "type": "reminder",
+            "message": message,
+            "scheduled_at": datetime.utcnow().isoformat(),
+            "trigger_at_minutes": delay_minutes,
+            "priority": priority
+        }
+
+        # Add to learning queue as placeholder
+        state._state.setdefault("reminders", []).append(reminder)
+        state._save_state(state._state)  # Pass state dict as argument
+
+        logger.info(f"[ACTIONS] Reminder scheduled: {message} (in {delay_minutes}min)")
+
+        return {
+            "message": message,
+            "delay_minutes": delay_minutes,
+            "note": "Reminder stored in self-state (scheduler integration pending)"
+        }
+
+    async def _summarize_session(
+        self,
+        parameters: Dict[str, Any],
+        context: Dict[str, Any]
+    ) -> Dict[str, Any]:
+        """
+        Generate a summary of current session.
+
+        Parameters:
+            - max_length: Max summary length in words
+            - focus_topics: Optional list of topics to emphasize
+        """
+        max_length = parameters.get("max_length", 200)
+        session_id = context.get("session_id", "unknown")
+
+        # Import summarizer (from deferred_summary or create simple one)
+        try:
+            from utils.deferred_summary import summarize_conversation
+
+            summary = await summarize_conversation(
+                session_id=session_id,
+                max_words=max_length
+            )
+
+            return {
+                "summary": summary,
+                "word_count": len(summary.split())
+            }
+
+        except ImportError:
+            # Fallback: simple summary
+            message_count = context.get("message_count", 0)
+            focus = context.get("monologue", {}).get("intent", "general")
+
+            summary = f"Session {session_id}: {message_count} messages exchanged, focused on {focus}."
+
+            return {
+                "summary": summary,
+                "word_count": len(summary.split()),
+                "note": "Simple summary (full summarizer not available)"
+            }
+
+    async def _learn_topic(
+        self,
+        parameters: Dict[str, Any],
+        context: Dict[str, Any]
+    ) -> Dict[str, Any]:
+        """
+        Add topic to learning queue.
+
+        Parameters:
+            - topic: Topic name (required)
+            - reason: Why this topic
+            - priority: 0.0-1.0 priority score
+        """
+        topic = parameters.get("topic")
+        if not topic:
+            raise ValueError("Topic required")
+
+        reason = parameters.get("reason", "autonomous learning")
+        priority = parameters.get("priority", 0.5)
+
+        # Import self-state manager
+        from autonomy.self.state import get_self_state_instance
+
+        state = get_self_state_instance()
+        state.add_learning_goal(topic)  # Only pass topic parameter
+
+        logger.info(f"[ACTIONS] Added to learning queue: {topic} (reason: {reason})")
+
+        return {
+            "topic": topic,
+            "reason": reason,
+            "queue_position": len(state._state.get("learning_queue", []))
+        }
+
+    async def _update_focus(
+        self,
+        parameters: Dict[str, Any],
+        context: Dict[str, Any]
+    ) -> Dict[str, Any]:
+        """
+        Update current focus area.
+
+        Parameters:
+            - focus: New focus area (required)
+            - reason: Why this focus
+        """
+        focus = parameters.get("focus")
+        if not focus:
+            raise ValueError("Focus required")
+
+        reason = parameters.get("reason", "autonomous update")
+
+        # Import self-state manager
+        from autonomy.self.state import get_self_state_instance
+
+        state = get_self_state_instance()
+        old_focus = state._state.get("focus", "none")
+
+        state._state["focus"] = focus
+        state._state["focus_updated_at"] = datetime.utcnow().isoformat()
+        state._state["focus_reason"] = reason
+        state._save_state(state._state)  # Pass state dict as argument
+
+        logger.info(f"[ACTIONS] Focus updated: {old_focus} -> {focus}")
+
+        return {
+            "old_focus": old_focus,
+            "new_focus": focus,
+            "reason": reason
+        }
+
+    # ========================================
+    # Utility Methods
+    # ========================================
+
+    def get_allowed_actions(self) -> List[str]:
+        """Get list of all allowed action types."""
+        return list(self.allowed_actions.keys())
+
+    def get_action_log(self, limit: int = 50) -> List[Dict[str, Any]]:
+        """
+        Get recent action log.
+
+        Args:
+            limit: Max number of entries to return
+
+        Returns:
+            List of action records
+        """
+        return self.action_log[-limit:]
+
+    def clear_action_log(self) -> None:
+        """Clear action log."""
+        self.action_log = []
+        logger.info("[ACTIONS] Action log cleared")
+
+    def validate_action(self, action_type: str, parameters: Dict[str, Any]) -> Dict[str, Any]:
+        """
+        Validate an action without executing it.
+
+        Args:
+            action_type: Type of action
+            parameters: Action parameters
+
+        Returns:
+            {
+                "valid": bool,
+                "action": action_type,
+                "errors": [error messages] or []
+            }
+        """
+        errors = []
+
+        # Check whitelist
+        if action_type not in self.allowed_actions:
+            errors.append(f"Action '{action_type}' not in whitelist")
+
+        # Check required parameters (basic validation)
+        if action_type == "create_memory" and not parameters.get("text"):
+            errors.append("Memory 'text' parameter required")
+
+        if action_type == "update_goal" and not parameters.get("goal_id"):
+            errors.append("Goal 'goal_id' parameter required")
+
+        if action_type == "schedule_reminder" and not parameters.get("message"):
+            errors.append("Reminder 'message' parameter required")
+
+        if action_type == "learn_topic" and not parameters.get("topic"):
+            errors.append("Learning 'topic' parameter required")
+
+        if action_type == "update_focus" and not parameters.get("focus"):
+            errors.append("Focus 'focus' parameter required")
+
+        return {
+            "valid": len(errors) == 0,
+            "action": action_type,
+            "errors": errors
+        }
+
+
+# Singleton instance
+_action_manager_instance = None
+
+
+def get_action_manager() -> AutonomousActionManager:
+    """
+    Get singleton action manager instance.
+
+    Returns:
+        AutonomousActionManager instance
+    """
+    global _action_manager_instance
+    if _action_manager_instance is None:
+        _action_manager_instance = AutonomousActionManager()
+    return _action_manager_instance

cortex/autonomy/executive/__init__.py

@@ -0,0 +1 @@
+"""Executive planning and decision-making module."""

cortex/autonomy/executive/planner.py

@@ -0,0 +1,121 @@
+"""
+Executive planner - generates execution plans for complex requests.
+Activated when inner monologue sets consult_executive=true.
+"""
+
+import os
+import logging
+from typing import Dict, Any, Optional
+from llm.llm_router import call_llm
+
+EXECUTIVE_LLM = os.getenv("EXECUTIVE_LLM", "CLOUD").upper()
+VERBOSE_DEBUG = os.getenv("VERBOSE_DEBUG", "false").lower() == "true"
+
+logger = logging.getLogger(__name__)
+
+if VERBOSE_DEBUG:
+    logger.setLevel(logging.DEBUG)
+
+
+EXECUTIVE_SYSTEM_PROMPT = """
+You are Lyra's executive planning system.
+You create structured execution plans for complex tasks.
+You do NOT generate the final response - only the plan.
+
+Your plan should include:
+1. Task decomposition (break into steps)
+2. Required tools/resources
+3. Reasoning strategy
+4. Success criteria
+
+Return a concise plan in natural language.
+"""
+
+
+async def plan_execution(
+    user_prompt: str,
+    intent: str,
+    context_state: Dict[str, Any],
+    identity_block: Dict[str, Any]
+) -> Dict[str, Any]:
+    """
+    Generate execution plan for complex request.
+
+    Args:
+        user_prompt: User's message
+        intent: Detected intent from inner monologue
+        context_state: Full context
+        identity_block: Lyra's identity
+
+    Returns:
+        Plan dictionary with structure:
+        {
+            "summary": "One-line plan summary",
+            "plan_text": "Detailed plan",
+            "steps": ["step1", "step2", ...],
+            "tools_needed": ["RAG", "WEB", ...],
+            "estimated_complexity": "low | medium | high"
+        }
+    """
+
+    # Build planning prompt
+    tools_available = context_state.get("tools_available", [])
+
+    prompt = f"""{EXECUTIVE_SYSTEM_PROMPT}
+
+User request: {user_prompt}
+
+Detected intent: {intent}
+
+Available tools: {", ".join(tools_available) if tools_available else "None"}
+
+Session context:
+- Message count: {context_state.get('message_count', 0)}
+- Time since last message: {context_state.get('minutes_since_last_msg', 0):.1f} minutes
+- Active project: {context_state.get('active_project', 'None')}
+
+Generate a structured execution plan.
+"""
+
+    if VERBOSE_DEBUG:
+        logger.debug(f"[EXECUTIVE] Planning prompt:\n{prompt}")
+
+    # Call executive LLM
+    plan_text = await call_llm(
+        prompt,
+        backend=EXECUTIVE_LLM,
+        temperature=0.3,  # Lower temperature for planning
+        max_tokens=500
+    )
+
+    if VERBOSE_DEBUG:
+        logger.debug(f"[EXECUTIVE] Generated plan:\n{plan_text}")
+
+    # Parse plan (simple heuristic extraction for Phase 1)
+    steps = []
+    tools_needed = []
+
+    for line in plan_text.split('\n'):
+        line_lower = line.lower()
+        if any(marker in line_lower for marker in ['step', '1.', '2.', '3.', '-']):
+            steps.append(line.strip())
+
+        if tools_available:
+            for tool in tools_available:
+                if tool.lower() in line_lower and tool not in tools_needed:
+                    tools_needed.append(tool)
+
+    # Estimate complexity (simple heuristic)
+    complexity = "low"
+    if len(steps) > 3 or len(tools_needed) > 1:
+        complexity = "medium"
+    if len(steps) > 5 or "research" in intent.lower() or "analyze" in intent.lower():
+        complexity = "high"
+
+    return {
+        "summary": plan_text.split('\n')[0][:100] if plan_text else "Complex task execution plan",
+        "plan_text": plan_text,
+        "steps": steps[:10],  # Limit to 10 steps
+        "tools_needed": tools_needed,
+        "estimated_complexity": complexity
+    }

cortex/autonomy/learning/__init__.py

@@ -0,0 +1 @@
+"""Pattern learning and adaptation system."""

cortex/autonomy/learning/pattern_learner.py

@@ -0,0 +1,383 @@
+"""
+Pattern Learning System - learns from interaction patterns to improve autonomy.
+"""
+
+import logging
+import json
+import os
+from typing import Dict, List, Any, Optional
+from datetime import datetime
+from collections import defaultdict
+
+logger = logging.getLogger(__name__)
+
+
+class PatternLearner:
+    """
+    Learns from interaction patterns to improve Lyra's autonomous behavior.
+
+    Tracks:
+    - Topic frequencies (what users talk about)
+    - Time-of-day patterns (when users interact)
+    - User preferences (how users like responses)
+    - Successful response strategies (what works well)
+    """
+
+    def __init__(self, patterns_file: str = "/app/data/learned_patterns.json"):
+        """
+        Initialize pattern learner.
+
+        Args:
+            patterns_file: Path to persistent patterns storage
+        """
+        self.patterns_file = patterns_file
+        self.patterns = self._load_patterns()
+
+    def _load_patterns(self) -> Dict[str, Any]:
+        """Load patterns from disk."""
+        if os.path.exists(self.patterns_file):
+            try:
+                with open(self.patterns_file, 'r') as f:
+                    patterns = json.load(f)
+                    logger.info(f"[PATTERN_LEARNER] Loaded patterns from {self.patterns_file}")
+                    return patterns
+            except Exception as e:
+                logger.error(f"[PATTERN_LEARNER] Failed to load patterns: {e}")
+
+        # Initialize empty patterns
+        return {
+            "topic_frequencies": {},
+            "time_patterns": {},
+            "user_preferences": {},
+            "successful_strategies": {},
+            "interaction_count": 0,
+            "last_updated": datetime.utcnow().isoformat()
+        }
+
+    def _save_patterns(self) -> None:
+        """Save patterns to disk."""
+        try:
+            # Ensure directory exists
+            os.makedirs(os.path.dirname(self.patterns_file), exist_ok=True)
+
+            self.patterns["last_updated"] = datetime.utcnow().isoformat()
+
+            with open(self.patterns_file, 'w') as f:
+                json.dump(self.patterns, f, indent=2)
+
+            logger.debug(f"[PATTERN_LEARNER] Saved patterns to {self.patterns_file}")
+
+        except Exception as e:
+            logger.error(f"[PATTERN_LEARNER] Failed to save patterns: {e}")
+
+    async def learn_from_interaction(
+        self,
+        user_prompt: str,
+        response: str,
+        monologue: Dict[str, Any],
+        context: Dict[str, Any]
+    ) -> None:
+        """
+        Learn from a single interaction.
+
+        Args:
+            user_prompt: User's message
+            response: Lyra's response
+            monologue: Inner monologue analysis
+            context: Full context state
+        """
+        self.patterns["interaction_count"] += 1
+
+        # Learn topic frequencies
+        self._learn_topics(user_prompt, monologue)
+
+        # Learn time patterns
+        self._learn_time_patterns()
+
+        # Learn user preferences
+        self._learn_preferences(monologue, context)
+
+        # Learn successful strategies
+        self._learn_strategies(monologue, response, context)
+
+        # Save periodically (every 10 interactions)
+        if self.patterns["interaction_count"] % 10 == 0:
+            self._save_patterns()
+
+    def _learn_topics(self, user_prompt: str, monologue: Dict[str, Any]) -> None:
+        """Track topic frequencies."""
+        intent = monologue.get("intent", "unknown")
+
+        # Increment topic counter
+        topic_freq = self.patterns["topic_frequencies"]
+        topic_freq[intent] = topic_freq.get(intent, 0) + 1
+
+        # Extract keywords (simple approach - words > 5 chars)
+        keywords = [word.lower() for word in user_prompt.split() if len(word) > 5]
+
+        for keyword in keywords:
+            topic_freq[f"keyword:{keyword}"] = topic_freq.get(f"keyword:{keyword}", 0) + 1
+
+        logger.debug(f"[PATTERN_LEARNER] Topic learned: {intent}")
+
+    def _learn_time_patterns(self) -> None:
+        """Track time-of-day patterns."""
+        now = datetime.utcnow()
+        hour = now.hour
+
+        # Track interactions by hour
+        time_patterns = self.patterns["time_patterns"]
+        hour_key = f"hour_{hour:02d}"
+        time_patterns[hour_key] = time_patterns.get(hour_key, 0) + 1
+
+        # Track day of week
+        day_key = f"day_{now.strftime('%A').lower()}"
+        time_patterns[day_key] = time_patterns.get(day_key, 0) + 1
+
+    def _learn_preferences(self, monologue: Dict[str, Any], context: Dict[str, Any]) -> None:
+        """Learn user preferences from detected tone and depth."""
+        tone = monologue.get("tone", "neutral")
+        depth = monologue.get("depth", "medium")
+
+        prefs = self.patterns["user_preferences"]
+
+        # Track preferred tone
+        prefs.setdefault("tone_counts", {})
+        prefs["tone_counts"][tone] = prefs["tone_counts"].get(tone, 0) + 1
+
+        # Track preferred depth
+        prefs.setdefault("depth_counts", {})
+        prefs["depth_counts"][depth] = prefs["depth_counts"].get(depth, 0) + 1
+
+    def _learn_strategies(
+        self,
+        monologue: Dict[str, Any],
+        response: str,
+        context: Dict[str, Any]
+    ) -> None:
+        """
+        Learn which response strategies are successful.
+
+        Success indicators:
+        - Executive was consulted and plan generated
+        - Response length matches depth request
+        - Tone matches request
+        """
+        intent = monologue.get("intent", "unknown")
+        executive_used = context.get("executive_plan") is not None
+
+        strategies = self.patterns["successful_strategies"]
+        strategies.setdefault(intent, {})
+
+        # Track executive usage for this intent
+        if executive_used:
+            key = f"{intent}:executive_used"
+            strategies.setdefault(key, 0)
+            strategies[key] += 1
+
+        # Track response length patterns
+        response_length = len(response.split())
+        depth = monologue.get("depth", "medium")
+
+        length_key = f"{depth}:avg_words"
+        if length_key not in strategies:
+            strategies[length_key] = response_length
+        else:
+            # Running average
+            strategies[length_key] = (strategies[length_key] + response_length) / 2
+
+    # ========================================
+    # Pattern Analysis and Recommendations
+    # ========================================
+
+    def get_top_topics(self, limit: int = 10) -> List[tuple]:
+        """
+        Get most frequent topics.
+
+        Args:
+            limit: Max number of topics to return
+
+        Returns:
+            List of (topic, count) tuples, sorted by count
+        """
+        topics = self.patterns["topic_frequencies"]
+        sorted_topics = sorted(topics.items(), key=lambda x: x[1], reverse=True)
+        return sorted_topics[:limit]
+
+    def get_preferred_tone(self) -> str:
+        """
+        Get user's most preferred tone.
+
+        Returns:
+            Preferred tone string
+        """
+        prefs = self.patterns["user_preferences"]
+        tone_counts = prefs.get("tone_counts", {})
+
+        if not tone_counts:
+            return "neutral"
+
+        return max(tone_counts.items(), key=lambda x: x[1])[0]
+
+    def get_preferred_depth(self) -> str:
+        """
+        Get user's most preferred response depth.
+
+        Returns:
+            Preferred depth string
+        """
+        prefs = self.patterns["user_preferences"]
+        depth_counts = prefs.get("depth_counts", {})
+
+        if not depth_counts:
+            return "medium"
+
+        return max(depth_counts.items(), key=lambda x: x[1])[0]
+
+    def get_peak_hours(self, limit: int = 3) -> List[int]:
+        """
+        Get peak interaction hours.
+
+        Args:
+            limit: Number of top hours to return
+
+        Returns:
+            List of hours (0-23)
+        """
+        time_patterns = self.patterns["time_patterns"]
+        hour_counts = {k: v for k, v in time_patterns.items() if k.startswith("hour_")}
+
+        if not hour_counts:
+            return []
+
+        sorted_hours = sorted(hour_counts.items(), key=lambda x: x[1], reverse=True)
+        top_hours = sorted_hours[:limit]
+
+        # Extract hour numbers
+        return [int(h[0].split("_")[1]) for h in top_hours]
+
+    def should_use_executive(self, intent: str) -> bool:
+        """
+        Recommend whether to use executive for given intent based on patterns.
+
+        Args:
+            intent: Intent type
+
+        Returns:
+            True if executive is recommended
+        """
+        strategies = self.patterns["successful_strategies"]
+        key = f"{intent}:executive_used"
+
+        # If we've used executive for this intent >= 3 times, recommend it
+        return strategies.get(key, 0) >= 3
+
+    def get_recommended_response_length(self, depth: str) -> int:
+        """
+        Get recommended response length in words for given depth.
+
+        Args:
+            depth: Depth level (short/medium/deep)
+
+        Returns:
+            Recommended word count
+        """
+        strategies = self.patterns["successful_strategies"]
+        key = f"{depth}:avg_words"
+
+        avg_length = strategies.get(key, None)
+
+        if avg_length:
+            return int(avg_length)
+
+        # Defaults if no pattern learned
+        defaults = {
+            "short": 50,
+            "medium": 150,
+            "deep": 300
+        }
+
+        return defaults.get(depth, 150)
+
+    def get_insights(self) -> Dict[str, Any]:
+        """
+        Get high-level insights from learned patterns.
+
+        Returns:
+            {
+                "total_interactions": int,
+                "top_topics": [(topic, count), ...],
+                "preferred_tone": str,
+                "preferred_depth": str,
+                "peak_hours": [hours],
+                "learning_recommendations": [str]
+            }
+        """
+        recommendations = []
+
+        # Check if user consistently prefers certain settings
+        preferred_tone = self.get_preferred_tone()
+        preferred_depth = self.get_preferred_depth()
+
+        if preferred_tone != "neutral":
+            recommendations.append(f"User prefers {preferred_tone} tone")
+
+        if preferred_depth != "medium":
+            recommendations.append(f"User prefers {preferred_depth} depth responses")
+
+        # Check for recurring topics
+        top_topics = self.get_top_topics(limit=3)
+        if top_topics:
+            top_topic = top_topics[0][0]
+            recommendations.append(f"Consider adding '{top_topic}' to learning queue")
+
+        return {
+            "total_interactions": self.patterns["interaction_count"],
+            "top_topics": self.get_top_topics(limit=5),
+            "preferred_tone": preferred_tone,
+            "preferred_depth": preferred_depth,
+            "peak_hours": self.get_peak_hours(limit=3),
+            "learning_recommendations": recommendations
+        }
+
+    def reset_patterns(self) -> None:
+        """Reset all learned patterns (use with caution)."""
+        self.patterns = {
+            "topic_frequencies": {},
+            "time_patterns": {},
+            "user_preferences": {},
+            "successful_strategies": {},
+            "interaction_count": 0,
+            "last_updated": datetime.utcnow().isoformat()
+        }
+        self._save_patterns()
+        logger.warning("[PATTERN_LEARNER] Patterns reset")
+
+    def export_patterns(self) -> Dict[str, Any]:
+        """
+        Export all patterns for analysis.
+
+        Returns:
+            Complete patterns dict
+        """
+        return self.patterns.copy()
+
+
+# Singleton instance
+_learner_instance = None
+
+
+def get_pattern_learner(patterns_file: str = "/app/data/learned_patterns.json") -> PatternLearner:
+    """
+    Get singleton pattern learner instance.
+
+    Args:
+        patterns_file: Path to patterns file (only used on first call)
+
+    Returns:
+        PatternLearner instance
+    """
+    global _learner_instance
+    if _learner_instance is None:
+        _learner_instance = PatternLearner(patterns_file=patterns_file)
+    return _learner_instance

cortex/autonomy/monologue/__init__.py

@@ -0,0 +1 @@
+# Inner monologue module

cortex/autonomy/monologue/monologue.py

@@ -0,0 +1,115 @@
+import os
+import json
+import logging
+from typing import Dict
+from llm.llm_router import call_llm
+
+# Configuration
+MONOLOGUE_LLM = os.getenv("MONOLOGUE_LLM", "PRIMARY").upper()
+VERBOSE_DEBUG = os.getenv("VERBOSE_DEBUG", "false").lower() == "true"
+
+# Logger
+logger = logging.getLogger(__name__)
+
+if VERBOSE_DEBUG:
+    logger.setLevel(logging.DEBUG)
+    console_handler = logging.StreamHandler()
+    console_handler.setFormatter(logging.Formatter(
+        '%(asctime)s [MONOLOGUE] %(levelname)s: %(message)s',
+        datefmt='%H:%M:%S'
+    ))
+    logger.addHandler(console_handler)
+
+MONOLOGUE_SYSTEM_PROMPT = """
+You are Lyra's inner monologue.
+You think privately.
+You do NOT speak to the user.
+You do NOT solve the task.
+You only reflect on intent, tone, and depth.
+
+Return ONLY valid JSON with:
+- intent (string)
+- tone (neutral | warm | focused | playful | direct)
+- depth (short | medium | deep)
+- consult_executive (true | false)
+"""
+
+class InnerMonologue:
+    async def process(self, context: Dict) -> Dict:
+        # Build full prompt with system instructions merged in
+        full_prompt = f"""{MONOLOGUE_SYSTEM_PROMPT}
+
+User message:
+{context['user_message']}
+
+Self state:
+{context['self_state']}
+
+Context summary:
+{context['context_summary']}
+
+Output JSON only:
+"""
+
+        # Call LLM using configured backend
+        if VERBOSE_DEBUG:
+            logger.debug(f"[InnerMonologue] Calling LLM with backend: {MONOLOGUE_LLM}")
+            logger.debug(f"[InnerMonologue] Prompt length: {len(full_prompt)} chars")
+
+        result = await call_llm(
+            full_prompt,
+            backend=MONOLOGUE_LLM,
+            temperature=0.7,
+            max_tokens=200
+        )
+
+        if VERBOSE_DEBUG:
+            logger.debug(f"[InnerMonologue] Raw LLM response:")
+            logger.debug(f"{'='*80}")
+            logger.debug(result)
+            logger.debug(f"{'='*80}")
+            logger.debug(f"[InnerMonologue] Response length: {len(result) if result else 0} chars")
+
+        # Parse JSON response - extract just the JSON part if there's extra text
+        try:
+            # Try direct parsing first
+            parsed = json.loads(result)
+            if VERBOSE_DEBUG:
+                logger.debug(f"[InnerMonologue] Successfully parsed JSON directly: {parsed}")
+            return parsed
+        except json.JSONDecodeError:
+            # If direct parsing fails, try to extract JSON from the response
+            if VERBOSE_DEBUG:
+                logger.debug(f"[InnerMonologue] Direct JSON parse failed, attempting extraction...")
+
+            # Look for JSON object (starts with { and ends with })
+            import re
+            json_match = re.search(r'\{[^{}]*(?:\{[^{}]*\}[^{}]*)*\}', result, re.DOTALL)
+
+            if json_match:
+                json_str = json_match.group(0)
+                try:
+                    parsed = json.loads(json_str)
+                    if VERBOSE_DEBUG:
+                        logger.debug(f"[InnerMonologue] Successfully extracted and parsed JSON: {parsed}")
+                    return parsed
+                except json.JSONDecodeError as e:
+                    if VERBOSE_DEBUG:
+                        logger.warning(f"[InnerMonologue] Extracted JSON still invalid: {e}")
+            else:
+                if VERBOSE_DEBUG:
+                    logger.warning(f"[InnerMonologue] No JSON object found in response")
+
+            # Final fallback
+            if VERBOSE_DEBUG:
+                logger.warning(f"[InnerMonologue] All parsing attempts failed, using fallback")
+            else:
+                print(f"[InnerMonologue] JSON extraction failed")
+                print(f"[InnerMonologue] Raw response was: {result[:500]}")
+
+            return {
+                "intent": "unknown",
+                "tone": "neutral",
+                "depth": "medium",
+                "consult_executive": False
+            }

cortex/autonomy/proactive/__init__.py

@@ -0,0 +1 @@
+"""Proactive monitoring and suggestion system."""

cortex/autonomy/proactive/monitor.py

@@ -0,0 +1,321 @@
+"""
+Proactive Context Monitor - detects opportunities for autonomous suggestions.
+"""
+
+import logging
+import time
+from typing import Dict, List, Any, Optional
+from datetime import datetime, timedelta
+
+logger = logging.getLogger(__name__)
+
+
+class ProactiveMonitor:
+    """
+    Monitors conversation context and detects opportunities for proactive suggestions.
+
+    Triggers:
+    - Long silence → Check-in
+    - Learning queue + high curiosity → Suggest exploration
+    - Active goals → Progress reminders
+    - Conversation milestones → Offer summary
+    - Pattern detection → Helpful suggestions
+    """
+
+    def __init__(self, min_priority: float = 0.6):
+        """
+        Initialize proactive monitor.
+
+        Args:
+            min_priority: Minimum priority for suggestions (0.0-1.0)
+        """
+        self.min_priority = min_priority
+        self.last_suggestion_time = {}  # session_id -> timestamp
+        self.cooldown_seconds = 300  # 5 minutes between proactive suggestions
+
+    async def analyze_session(
+        self,
+        session_id: str,
+        context_state: Dict[str, Any],
+        self_state: Dict[str, Any]
+    ) -> Optional[Dict[str, Any]]:
+        """
+        Analyze session for proactive suggestion opportunities.
+
+        Args:
+            session_id: Current session ID
+            context_state: Full context including message history
+            self_state: Lyra's current self-state
+
+        Returns:
+            {
+                "suggestion": "text to append to response",
+                "priority": 0.0-1.0,
+                "reason": "why this suggestion",
+                "type": "check_in | learning | goal_reminder | summary | pattern"
+            }
+            or None if no suggestion
+        """
+        # Check cooldown
+        if not self._check_cooldown(session_id):
+            logger.debug(f"[PROACTIVE] Session {session_id} in cooldown, skipping")
+            return None
+
+        suggestions = []
+
+        # Check 1: Long silence detection
+        silence_suggestion = self._check_long_silence(context_state)
+        if silence_suggestion:
+            suggestions.append(silence_suggestion)
+
+        # Check 2: Learning queue + high curiosity
+        learning_suggestion = self._check_learning_opportunity(self_state)
+        if learning_suggestion:
+            suggestions.append(learning_suggestion)
+
+        # Check 3: Active goals reminder
+        goal_suggestion = self._check_active_goals(self_state, context_state)
+        if goal_suggestion:
+            suggestions.append(goal_suggestion)
+
+        # Check 4: Conversation milestones
+        milestone_suggestion = self._check_conversation_milestone(context_state)
+        if milestone_suggestion:
+            suggestions.append(milestone_suggestion)
+
+        # Check 5: Pattern-based suggestions
+        pattern_suggestion = self._check_patterns(context_state, self_state)
+        if pattern_suggestion:
+            suggestions.append(pattern_suggestion)
+
+        # Filter by priority and return highest
+        valid_suggestions = [s for s in suggestions if s["priority"] >= self.min_priority]
+
+        if not valid_suggestions:
+            return None
+
+        # Return highest priority suggestion
+        best_suggestion = max(valid_suggestions, key=lambda x: x["priority"])
+
+        # Update cooldown timer
+        self._update_cooldown(session_id)
+
+        logger.info(f"[PROACTIVE] Suggestion generated: {best_suggestion['type']} (priority: {best_suggestion['priority']:.2f})")
+
+        return best_suggestion
+
+    def _check_cooldown(self, session_id: str) -> bool:
+        """Check if session is past cooldown period."""
+        if session_id not in self.last_suggestion_time:
+            return True
+
+        elapsed = time.time() - self.last_suggestion_time[session_id]
+        return elapsed >= self.cooldown_seconds
+
+    def _update_cooldown(self, session_id: str) -> None:
+        """Update cooldown timer for session."""
+        self.last_suggestion_time[session_id] = time.time()
+
+    def _check_long_silence(self, context_state: Dict[str, Any]) -> Optional[Dict[str, Any]]:
+        """
+        Check if user has been silent for a long time.
+        """
+        minutes_since_last = context_state.get("minutes_since_last_msg", 0)
+
+        # If > 30 minutes, suggest check-in
+        if minutes_since_last > 30:
+            return {
+                "suggestion": "\n\n[Aside: I'm still here if you need anything!]",
+                "priority": 0.7,
+                "reason": f"User silent for {minutes_since_last:.0f} minutes",
+                "type": "check_in"
+            }
+
+        return None
+
+    def _check_learning_opportunity(self, self_state: Dict[str, Any]) -> Optional[Dict[str, Any]]:
+        """
+        Check if Lyra has learning queue items and high curiosity.
+        """
+        learning_queue = self_state.get("learning_queue", [])
+        curiosity = self_state.get("curiosity", 0.5)
+
+        # If curiosity > 0.7 and learning queue exists
+        if curiosity > 0.7 and learning_queue:
+            topic = learning_queue[0] if learning_queue else "new topics"
+            return {
+                "suggestion": f"\n\n[Aside: I've been curious about {topic} lately. Would you like to explore it together?]",
+                "priority": 0.65,
+                "reason": f"High curiosity ({curiosity:.2f}) and learning queue present",
+                "type": "learning"
+            }
+
+        return None
+
+    def _check_active_goals(
+        self,
+        self_state: Dict[str, Any],
+        context_state: Dict[str, Any]
+    ) -> Optional[Dict[str, Any]]:
+        """
+        Check if there are active goals worth reminding about.
+        """
+        active_goals = self_state.get("active_goals", [])
+
+        if not active_goals:
+            return None
+
+        # Check if we've had multiple messages without goal progress
+        message_count = context_state.get("message_count", 0)
+
+        # Every 10 messages, consider goal reminder
+        if message_count % 10 == 0 and message_count > 0:
+            goal = active_goals[0]  # First active goal
+            goal_name = goal if isinstance(goal, str) else goal.get("name", "your goal")
+
+            return {
+                "suggestion": f"\n\n[Aside: Still thinking about {goal_name}. Let me know if you want to work on it.]",
+                "priority": 0.6,
+                "reason": f"Active goal present, {message_count} messages since start",
+                "type": "goal_reminder"
+            }
+
+        return None
+
+    def _check_conversation_milestone(self, context_state: Dict[str, Any]) -> Optional[Dict[str, Any]]:
+        """
+        Check for conversation milestones (e.g., every 50 messages).
+        """
+        message_count = context_state.get("message_count", 0)
+
+        # Every 50 messages, offer summary
+        if message_count > 0 and message_count % 50 == 0:
+            return {
+                "suggestion": f"\n\n[Aside: We've exchanged {message_count} messages! Would you like a summary of our conversation?]",
+                "priority": 0.65,
+                "reason": f"Milestone: {message_count} messages",
+                "type": "summary"
+            }
+
+        return None
+
+    def _check_patterns(
+        self,
+        context_state: Dict[str, Any],
+        self_state: Dict[str, Any]
+    ) -> Optional[Dict[str, Any]]:
+        """
+        Check for behavioral patterns that merit suggestions.
+        """
+        # Get current focus
+        focus = self_state.get("focus", "")
+
+        # Check if user keeps asking similar questions (detected via focus)
+        if focus and "repeated" in focus.lower():
+            return {
+                "suggestion": "\n\n[Aside: I notice we keep coming back to this topic. Would it help to create a summary or action plan?]",
+                "priority": 0.7,
+                "reason": "Repeated topic detected",
+                "type": "pattern"
+            }
+
+        # Check energy levels - if Lyra is low energy, maybe suggest break
+        energy = self_state.get("energy", 0.8)
+        if energy < 0.3:
+            return {
+                "suggestion": "\n\n[Aside: We've been at this for a while. Need a break or want to keep going?]",
+                "priority": 0.65,
+                "reason": f"Low energy ({energy:.2f})",
+                "type": "pattern"
+            }
+
+        return None
+
+    def format_suggestion(self, suggestion: Dict[str, Any]) -> str:
+        """
+        Format suggestion for appending to response.
+
+        Args:
+            suggestion: Suggestion dict from analyze_session()
+
+        Returns:
+            Formatted string to append to response
+        """
+        return suggestion.get("suggestion", "")
+
+    def set_cooldown_duration(self, seconds: int) -> None:
+        """
+        Update cooldown duration.
+
+        Args:
+            seconds: New cooldown duration
+        """
+        self.cooldown_seconds = seconds
+        logger.info(f"[PROACTIVE] Cooldown updated to {seconds}s")
+
+    def reset_cooldown(self, session_id: str) -> None:
+        """
+        Reset cooldown for a specific session.
+
+        Args:
+            session_id: Session to reset
+        """
+        if session_id in self.last_suggestion_time:
+            del self.last_suggestion_time[session_id]
+            logger.info(f"[PROACTIVE] Cooldown reset for session {session_id}")
+
+    def get_session_stats(self, session_id: str) -> Dict[str, Any]:
+        """
+        Get stats for a session's proactive monitoring.
+
+        Args:
+            session_id: Session to check
+
+        Returns:
+            {
+                "last_suggestion_time": timestamp or None,
+                "seconds_since_last": int,
+                "cooldown_active": bool,
+                "cooldown_remaining": int
+            }
+        """
+        last_time = self.last_suggestion_time.get(session_id)
+
+        if not last_time:
+            return {
+                "last_suggestion_time": None,
+                "seconds_since_last": 0,
+                "cooldown_active": False,
+                "cooldown_remaining": 0
+            }
+
+        seconds_since = int(time.time() - last_time)
+        cooldown_active = seconds_since < self.cooldown_seconds
+        cooldown_remaining = max(0, self.cooldown_seconds - seconds_since)
+
+        return {
+            "last_suggestion_time": last_time,
+            "seconds_since_last": seconds_since,
+            "cooldown_active": cooldown_active,
+            "cooldown_remaining": cooldown_remaining
+        }
+
+
+# Singleton instance
+_monitor_instance = None
+
+
+def get_proactive_monitor(min_priority: float = 0.6) -> ProactiveMonitor:
+    """
+    Get singleton proactive monitor instance.
+
+    Args:
+        min_priority: Minimum priority threshold (only used on first call)
+
+    Returns:
+        ProactiveMonitor instance
+    """
+    global _monitor_instance
+    if _monitor_instance is None:
+        _monitor_instance = ProactiveMonitor(min_priority=min_priority)
+    return _monitor_instance

cortex/autonomy/self/__init__.py

@@ -0,0 +1 @@
+# Self state module

cortex/autonomy/self/analyzer.py

@@ -0,0 +1,74 @@
+"""
+Analyze interactions and update self-state accordingly.
+"""
+
+import logging
+from typing import Dict, Any
+from .state import update_self_state
+
+logger = logging.getLogger(__name__)
+
+
+async def analyze_and_update_state(
+    monologue: Dict[str, Any],
+    user_prompt: str,
+    response: str,
+    context: Dict[str, Any]
+) -> None:
+    """
+    Analyze interaction and update self-state.
+
+    This runs after response generation to update Lyra's internal state
+    based on the interaction.
+
+    Args:
+        monologue: Inner monologue output
+        user_prompt: User's message
+        response: Lyra's response
+        context: Full context state
+    """
+
+    # Simple heuristics for state updates
+    # TODO: Replace with LLM-based sentiment analysis in Phase 2
+
+    mood_delta = 0.0
+    energy_delta = 0.0
+    confidence_delta = 0.0
+    curiosity_delta = 0.0
+    new_focus = None
+
+    # Analyze intent from monologue
+    intent = monologue.get("intent", "").lower() if monologue else ""
+
+    if "technical" in intent or "complex" in intent:
+        energy_delta = -0.05  # Deep thinking is tiring
+        confidence_delta = 0.05 if len(response) > 200 else -0.05
+        new_focus = "technical_problem"
+
+    elif "creative" in intent or "brainstorm" in intent:
+        mood_delta = 0.1  # Creative work is engaging
+        curiosity_delta = 0.1
+        new_focus = "creative_exploration"
+
+    elif "clarification" in intent or "confused" in intent:
+        confidence_delta = -0.05
+        new_focus = "understanding_user"
+
+    elif "simple" in intent or "casual" in intent:
+        energy_delta = 0.05  # Light conversation is refreshing
+        new_focus = "conversation"
+
+    # Check for learning opportunities (questions in user prompt)
+    if "?" in user_prompt and any(word in user_prompt.lower() for word in ["how", "why", "what"]):
+        curiosity_delta += 0.05
+
+    # Update state
+    update_self_state(
+        mood_delta=mood_delta,
+        energy_delta=energy_delta,
+        new_focus=new_focus,
+        confidence_delta=confidence_delta,
+        curiosity_delta=curiosity_delta
+    )
+
+    logger.info(f"Self-state updated based on interaction: focus={new_focus}")

cortex/autonomy/self/state.py

@@ -0,0 +1,189 @@
+"""
+Self-state management for Project Lyra.
+Maintains persistent identity, mood, energy, and focus across sessions.
+"""
+
+import json
+import logging
+import os
+from datetime import datetime
+from pathlib import Path
+from typing import Dict, Any, Optional
+
+# Configuration
+STATE_FILE = Path(os.getenv("SELF_STATE_FILE", "/app/data/self_state.json"))
+VERBOSE_DEBUG = os.getenv("VERBOSE_DEBUG", "false").lower() == "true"
+
+logger = logging.getLogger(__name__)
+
+if VERBOSE_DEBUG:
+    logger.setLevel(logging.DEBUG)
+
+# Default state structure
+DEFAULT_STATE = {
+    "mood": "neutral",
+    "energy": 0.8,
+    "focus": "user_request",
+    "confidence": 0.7,
+    "curiosity": 0.5,
+    "last_updated": None,
+    "interaction_count": 0,
+    "learning_queue": [],  # Topics Lyra wants to explore
+    "active_goals": [],  # Self-directed goals
+    "preferences": {
+        "verbosity": "medium",
+        "formality": "casual",
+        "proactivity": 0.3  # How likely to suggest things unprompted
+    },
+    "metadata": {
+        "version": "1.0",
+        "created_at": None
+    }
+}
+
+
+class SelfState:
+    """Manages Lyra's persistent self-state."""
+
+    def __init__(self):
+        self._state = self._load_state()
+
+    def _load_state(self) -> Dict[str, Any]:
+        """Load state from disk or create default."""
+        if STATE_FILE.exists():
+            try:
+                with open(STATE_FILE, 'r') as f:
+                    state = json.load(f)
+                    logger.info(f"Loaded self-state from {STATE_FILE}")
+                    return state
+            except Exception as e:
+                logger.error(f"Failed to load self-state: {e}")
+                return self._create_default_state()
+        else:
+            return self._create_default_state()
+
+    def _create_default_state(self) -> Dict[str, Any]:
+        """Create and save default state."""
+        state = DEFAULT_STATE.copy()
+        state["metadata"]["created_at"] = datetime.now().isoformat()
+        state["last_updated"] = datetime.now().isoformat()
+        self._save_state(state)
+        logger.info("Created new default self-state")
+        return state
+
+    def _save_state(self, state: Dict[str, Any]) -> None:
+        """Persist state to disk."""
+        try:
+            STATE_FILE.parent.mkdir(parents=True, exist_ok=True)
+            with open(STATE_FILE, 'w') as f:
+                json.dump(state, f, indent=2)
+            if VERBOSE_DEBUG:
+                logger.debug(f"Saved self-state to {STATE_FILE}")
+        except Exception as e:
+            logger.error(f"Failed to save self-state: {e}")
+
+    def get_state(self) -> Dict[str, Any]:
+        """Get current state snapshot."""
+        return self._state.copy()
+
+    def update_from_interaction(
+        self,
+        mood_delta: float = 0.0,
+        energy_delta: float = 0.0,
+        new_focus: Optional[str] = None,
+        confidence_delta: float = 0.0,
+        curiosity_delta: float = 0.0
+    ) -> None:
+        """
+        Update state based on interaction.
+
+        Args:
+            mood_delta: Change in mood (-1.0 to 1.0)
+            energy_delta: Change in energy (-1.0 to 1.0)
+            new_focus: New focus area
+            confidence_delta: Change in confidence
+            curiosity_delta: Change in curiosity
+        """
+        # Apply deltas with bounds checking
+        self._state["energy"] = max(0.0, min(1.0,
+            self._state.get("energy", 0.8) + energy_delta))
+
+        self._state["confidence"] = max(0.0, min(1.0,
+            self._state.get("confidence", 0.7) + confidence_delta))
+
+        self._state["curiosity"] = max(0.0, min(1.0,
+            self._state.get("curiosity", 0.5) + curiosity_delta))
+
+        # Update focus if provided
+        if new_focus:
+            self._state["focus"] = new_focus
+
+        # Update mood (simplified sentiment)
+        if mood_delta != 0:
+            mood_map = ["frustrated", "neutral", "engaged", "excited"]
+            current_mood_idx = 1  # neutral default
+            if self._state.get("mood") in mood_map:
+                current_mood_idx = mood_map.index(self._state["mood"])
+
+            new_mood_idx = max(0, min(len(mood_map) - 1,
+                int(current_mood_idx + mood_delta * 2)))
+            self._state["mood"] = mood_map[new_mood_idx]
+
+        # Increment interaction counter
+        self._state["interaction_count"] = self._state.get("interaction_count", 0) + 1
+        self._state["last_updated"] = datetime.now().isoformat()
+
+        # Persist changes
+        self._save_state(self._state)
+
+        if VERBOSE_DEBUG:
+            logger.debug(f"Updated self-state: mood={self._state['mood']}, "
+                        f"energy={self._state['energy']:.2f}, "
+                        f"confidence={self._state['confidence']:.2f}")
+
+    def add_learning_goal(self, topic: str) -> None:
+        """Add topic to learning queue."""
+        queue = self._state.get("learning_queue", [])
+        if topic not in [item.get("topic") for item in queue]:
+            queue.append({
+                "topic": topic,
+                "added_at": datetime.now().isoformat(),
+                "priority": 0.5
+            })
+            self._state["learning_queue"] = queue
+            self._save_state(self._state)
+            logger.info(f"Added learning goal: {topic}")
+
+    def add_active_goal(self, goal: str, context: str = "") -> None:
+        """Add self-directed goal."""
+        goals = self._state.get("active_goals", [])
+        goals.append({
+            "goal": goal,
+            "context": context,
+            "created_at": datetime.now().isoformat(),
+            "status": "active"
+        })
+        self._state["active_goals"] = goals
+        self._save_state(self._state)
+        logger.info(f"Added active goal: {goal}")
+
+
+# Global instance
+_self_state_instance = None
+
+def get_self_state_instance() -> SelfState:
+    """Get or create global SelfState instance."""
+    global _self_state_instance
+    if _self_state_instance is None:
+        _self_state_instance = SelfState()
+    return _self_state_instance
+
+
+def load_self_state() -> Dict[str, Any]:
+    """Load self state - public API for backwards compatibility."""
+    return get_self_state_instance().get_state()
+
+
+def update_self_state(**kwargs) -> None:
+    """Update self state - public API."""
+    get_self_state_instance().update_from_interaction(**kwargs)

cortex/autonomy/tools/__init__.py

@@ -0,0 +1 @@
+"""Autonomous tool invocation system."""

cortex/autonomy/tools/decision_engine.py

@@ -0,0 +1,124 @@
+"""
+Tool Decision Engine - decides which tools to invoke autonomously.
+"""
+
+import logging
+from typing import Dict, List, Any
+
+logger = logging.getLogger(__name__)
+
+
+class ToolDecisionEngine:
+    """Decides which tools to invoke based on context analysis."""
+
+    async def analyze_tool_needs(
+        self,
+        user_prompt: str,
+        monologue: Dict[str, Any],
+        context_state: Dict[str, Any],
+        available_tools: List[str]
+    ) -> Dict[str, Any]:
+        """
+        Analyze if tools should be invoked and which ones.
+
+        Args:
+            user_prompt: User's message
+            monologue: Inner monologue analysis
+            context_state: Full context
+            available_tools: List of available tools
+
+        Returns:
+            {
+                "should_invoke_tools": bool,
+                "tools_to_invoke": [
+                    {
+                        "tool": "RAG | WEB | WEATHER | etc",
+                        "query": "search query",
+                        "reason": "why this tool",
+                        "priority": 0.0-1.0
+                    },
+                    ...
+                ],
+                "confidence": 0.0-1.0
+            }
+        """
+
+        tools_to_invoke = []
+
+        # Check for memory/context needs
+        if any(word in user_prompt.lower() for word in [
+            "remember", "you said", "we discussed", "earlier", "before",
+            "last time", "previously", "what did"
+        ]):
+            tools_to_invoke.append({
+                "tool": "RAG",
+                "query": user_prompt,
+                "reason": "User references past conversation",
+                "priority": 0.9
+            })
+
+        # Check for web search needs
+        if any(word in user_prompt.lower() for word in [
+            "current", "latest", "news", "today", "what's happening",
+            "look up", "search for", "find information", "recent"
+        ]):
+            tools_to_invoke.append({
+                "tool": "WEB",
+                "query": user_prompt,
+                "reason": "Requires current information",
+                "priority": 0.8
+            })
+
+        # Check for weather needs
+        if any(word in user_prompt.lower() for word in [
+            "weather", "temperature", "forecast", "rain", "sunny", "climate"
+        ]):
+            tools_to_invoke.append({
+                "tool": "WEATHER",
+                "query": user_prompt,
+                "reason": "Weather information requested",
+                "priority": 0.95
+            })
+
+        # Check for code-related needs
+        if any(word in user_prompt.lower() for word in [
+            "code", "function", "debug", "implement", "algorithm",
+            "programming", "script", "syntax"
+        ]):
+            if "CODEBRAIN" in available_tools:
+                tools_to_invoke.append({
+                    "tool": "CODEBRAIN",
+                    "query": user_prompt,
+                    "reason": "Code-related task",
+                    "priority": 0.85
+                })
+
+        # Proactive RAG for complex queries (based on monologue)
+        intent = monologue.get("intent", "") if monologue else ""
+        if monologue and monologue.get("consult_executive"):
+            # Complex query - might benefit from context
+            if not any(t["tool"] == "RAG" for t in tools_to_invoke):
+                tools_to_invoke.append({
+                    "tool": "RAG",
+                    "query": user_prompt,
+                    "reason": "Complex query benefits from context",
+                    "priority": 0.6
+                })
+
+        # Sort by priority
+        tools_to_invoke.sort(key=lambda x: x["priority"], reverse=True)
+
+        max_priority = max([t["priority"] for t in tools_to_invoke]) if tools_to_invoke else 0.0
+
+        result = {
+            "should_invoke_tools": len(tools_to_invoke) > 0,
+            "tools_to_invoke": tools_to_invoke,
+            "confidence": max_priority
+        }
+
+        if tools_to_invoke:
+            logger.info(f"[TOOL_DECISION] Autonomous tool invocation recommended: {len(tools_to_invoke)} tools")
+            for tool in tools_to_invoke:
+                logger.info(f"  - {tool['tool']} (priority: {tool['priority']:.2f}): {tool['reason']}")
+
+        return result

cortex/autonomy/tools/orchestrator.py

@@ -0,0 +1,357 @@
+"""
+Tool Orchestrator - executes autonomous tool invocations asynchronously.
+"""
+
+import asyncio
+import logging
+from typing import Dict, List, Any, Optional
+import os
+
+logger = logging.getLogger(__name__)
+
+
+class ToolOrchestrator:
+    """Orchestrates async tool execution and result aggregation."""
+
+    def __init__(self, tool_timeout: int = 30):
+        """
+        Initialize orchestrator.
+
+        Args:
+            tool_timeout: Max seconds per tool call (default 30)
+        """
+        self.tool_timeout = tool_timeout
+        self.available_tools = self._discover_tools()
+
+    def _discover_tools(self) -> Dict[str, Any]:
+        """Discover available tool modules."""
+        tools = {}
+
+        # Import tool modules as they become available
+        if os.getenv("NEOMEM_ENABLED", "false").lower() == "true":
+            try:
+                from memory.neomem_client import search_neomem
+                tools["RAG"] = search_neomem
+                logger.debug("[ORCHESTRATOR] RAG tool available")
+            except ImportError:
+                logger.debug("[ORCHESTRATOR] RAG tool not available")
+        else:
+            logger.info("[ORCHESTRATOR] NEOMEM_ENABLED is false; RAG tool disabled")
+
+        try:
+            from integrations.web_search import web_search
+            tools["WEB"] = web_search
+            logger.debug("[ORCHESTRATOR] WEB tool available")
+        except ImportError:
+            logger.debug("[ORCHESTRATOR] WEB tool not available")
+
+        try:
+            from integrations.weather import get_weather
+            tools["WEATHER"] = get_weather
+            logger.debug("[ORCHESTRATOR] WEATHER tool available")
+        except ImportError:
+            logger.debug("[ORCHESTRATOR] WEATHER tool not available")
+
+        try:
+            from integrations.codebrain import query_codebrain
+            tools["CODEBRAIN"] = query_codebrain
+            logger.debug("[ORCHESTRATOR] CODEBRAIN tool available")
+        except ImportError:
+            logger.debug("[ORCHESTRATOR] CODEBRAIN tool not available")
+
+        return tools
+
+    async def execute_tools(
+        self,
+        tools_to_invoke: List[Dict[str, Any]],
+        context_state: Dict[str, Any]
+    ) -> Dict[str, Any]:
+        """
+        Execute multiple tools asynchronously.
+
+        Args:
+            tools_to_invoke: List of tool specs from decision engine
+                [{"tool": "RAG", "query": "...", "reason": "...", "priority": 0.9}, ...]
+            context_state: Full context for tool execution
+
+        Returns:
+            {
+                "results": {
+                    "RAG": {...},
+                    "WEB": {...},
+                    ...
+                },
+                "execution_summary": {
+                    "tools_invoked": ["RAG", "WEB"],
+                    "successful": ["RAG"],
+                    "failed": ["WEB"],
+                    "total_time_ms": 1234
+                }
+            }
+        """
+        import time
+        start_time = time.time()
+
+        logger.info(f"[ORCHESTRATOR] Executing {len(tools_to_invoke)} tools asynchronously")
+
+        # Create tasks for each tool
+        tasks = []
+        tool_names = []
+
+        for tool_spec in tools_to_invoke:
+            tool_name = tool_spec["tool"]
+            query = tool_spec["query"]
+
+            if tool_name in self.available_tools:
+                task = self._execute_single_tool(tool_name, query, context_state)
+                tasks.append(task)
+                tool_names.append(tool_name)
+                logger.debug(f"[ORCHESTRATOR] Queued {tool_name}: {query[:50]}...")
+            else:
+                logger.warning(f"[ORCHESTRATOR] Tool {tool_name} not available, skipping")
+
+        # Execute all tools concurrently with timeout
+        results = {}
+        successful = []
+        failed = []
+
+        if tasks:
+            try:
+                # Wait for all tasks with global timeout
+                completed = await asyncio.wait_for(
+                    asyncio.gather(*tasks, return_exceptions=True),
+                    timeout=self.tool_timeout
+                )
+
+                # Process results
+                for tool_name, result in zip(tool_names, completed):
+                    if isinstance(result, Exception):
+                        logger.error(f"[ORCHESTRATOR] {tool_name} failed: {result}")
+                        results[tool_name] = {"error": str(result), "success": False}
+                        failed.append(tool_name)
+                    else:
+                        logger.info(f"[ORCHESTRATOR] {tool_name} completed successfully")
+                        results[tool_name] = result
+                        successful.append(tool_name)
+
+            except asyncio.TimeoutError:
+                logger.error(f"[ORCHESTRATOR] Global timeout ({self.tool_timeout}s) exceeded")
+                for tool_name in tool_names:
+                    if tool_name not in results:
+                        results[tool_name] = {"error": "timeout", "success": False}
+                        failed.append(tool_name)
+
+        end_time = time.time()
+        total_time_ms = int((end_time - start_time) * 1000)
+
+        execution_summary = {
+            "tools_invoked": tool_names,
+            "successful": successful,
+            "failed": failed,
+            "total_time_ms": total_time_ms
+        }
+
+        logger.info(f"[ORCHESTRATOR] Execution complete: {len(successful)}/{len(tool_names)} successful in {total_time_ms}ms")
+
+        return {
+            "results": results,
+            "execution_summary": execution_summary
+        }
+
+    async def _execute_single_tool(
+        self,
+        tool_name: str,
+        query: str,
+        context_state: Dict[str, Any]
+    ) -> Dict[str, Any]:
+        """
+        Execute a single tool with error handling.
+
+        Args:
+            tool_name: Name of tool (RAG, WEB, etc.)
+            query: Query string for the tool
+            context_state: Context for tool execution
+
+        Returns:
+            Tool-specific result dict
+        """
+        tool_func = self.available_tools.get(tool_name)
+        if not tool_func:
+            raise ValueError(f"Tool {tool_name} not available")
+
+        try:
+            logger.debug(f"[ORCHESTRATOR] Invoking {tool_name}...")
+
+            # Different tools have different signatures - adapt as needed
+            if tool_name == "RAG":
+                result = await self._invoke_rag(tool_func, query, context_state)
+            elif tool_name == "WEB":
+                result = await self._invoke_web(tool_func, query)
+            elif tool_name == "WEATHER":
+                result = await self._invoke_weather(tool_func, query)
+            elif tool_name == "CODEBRAIN":
+                result = await self._invoke_codebrain(tool_func, query, context_state)
+            else:
+                # Generic invocation
+                result = await tool_func(query)
+
+            return {
+                "success": True,
+                "tool": tool_name,
+                "query": query,
+                "data": result
+            }
+
+        except Exception as e:
+            logger.error(f"[ORCHESTRATOR] {tool_name} execution failed: {e}")
+            raise
+
+    async def _invoke_rag(self, func, query: str, context: Dict[str, Any]) -> Any:
+        """Invoke RAG tool (NeoMem search)."""
+        session_id = context.get("session_id", "unknown")
+        # RAG searches memory for relevant past interactions
+        try:
+            results = await func(query, limit=5, session_id=session_id)
+            return results
+        except Exception as e:
+            logger.warning(f"[ORCHESTRATOR] RAG invocation failed, returning empty: {e}")
+            return []
+
+    async def _invoke_web(self, func, query: str) -> Any:
+        """Invoke web search tool."""
+        try:
+            results = await func(query, max_results=5)
+            return results
+        except Exception as e:
+            logger.warning(f"[ORCHESTRATOR] WEB invocation failed: {e}")
+            return {"error": str(e), "results": []}
+
+    async def _invoke_weather(self, func, query: str) -> Any:
+        """Invoke weather tool."""
+        # Extract location from query (simple heuristic)
+        # In future: use LLM to extract location
+        try:
+            location = self._extract_location(query)
+            results = await func(location)
+            return results
+        except Exception as e:
+            logger.warning(f"[ORCHESTRATOR] WEATHER invocation failed: {e}")
+            return {"error": str(e)}
+
+    async def _invoke_codebrain(self, func, query: str, context: Dict[str, Any]) -> Any:
+        """Invoke codebrain tool."""
+        try:
+            results = await func(query, context=context)
+            return results
+        except Exception as e:
+            logger.warning(f"[ORCHESTRATOR] CODEBRAIN invocation failed: {e}")
+            return {"error": str(e)}
+
+    def _extract_location(self, query: str) -> str:
+        """
+        Extract location from weather query.
+        Simple heuristic - in future use LLM.
+        """
+        # Common location indicators
+        indicators = ["in ", "at ", "for ", "weather in ", "temperature in "]
+
+        query_lower = query.lower()
+        for indicator in indicators:
+            if indicator in query_lower:
+                # Get text after indicator
+                parts = query_lower.split(indicator, 1)
+                if len(parts) > 1:
+                    location = parts[1].strip().split()[0]  # First word after indicator
+                    return location
+
+        # Default fallback
+        return "current location"
+
+    def format_results_for_context(self, orchestrator_result: Dict[str, Any]) -> str:
+        """
+        Format tool results for inclusion in context/prompt.
+
+        Args:
+            orchestrator_result: Output from execute_tools()
+
+        Returns:
+            Formatted string for prompt injection
+        """
+        results = orchestrator_result.get("results", {})
+        summary = orchestrator_result.get("execution_summary", {})
+
+        if not results:
+            return ""
+
+        formatted = "\n=== AUTONOMOUS TOOL RESULTS ===\n"
+
+        for tool_name, tool_result in results.items():
+            if tool_result.get("success", False):
+                formatted += f"\n[{tool_name}]\n"
+                data = tool_result.get("data", {})
+
+                # Format based on tool type
+                if tool_name == "RAG":
+                    formatted += self._format_rag_results(data)
+                elif tool_name == "WEB":
+                    formatted += self._format_web_results(data)
+                elif tool_name == "WEATHER":
+                    formatted += self._format_weather_results(data)
+                elif tool_name == "CODEBRAIN":
+                    formatted += self._format_codebrain_results(data)
+                else:
+                    formatted += f"{data}\n"
+            else:
+                formatted += f"\n[{tool_name}] - Failed: {tool_result.get('error', 'unknown')}\n"
+
+        formatted += f"\n(Tools executed in {summary.get('total_time_ms', 0)}ms)\n"
+        formatted += "=" * 40 + "\n"
+
+        return formatted
+
+    def _format_rag_results(self, data: Any) -> str:
+        """Format RAG/memory search results."""
+        if not data:
+            return "No relevant memories found.\n"
+
+        formatted = "Relevant memories:\n"
+        for i, item in enumerate(data[:3], 1):  # Top 3
+            text = item.get("text", item.get("content", str(item)))
+            formatted += f"  {i}. {text[:100]}...\n"
+        return formatted
+
+    def _format_web_results(self, data: Any) -> str:
+        """Format web search results."""
+        if isinstance(data, dict) and data.get("error"):
+            return f"Web search failed: {data['error']}\n"
+
+        results = data.get("results", []) if isinstance(data, dict) else data
+        if not results:
+            return "No web results found.\n"
+
+        formatted = "Web search results:\n"
+        for i, item in enumerate(results[:3], 1):  # Top 3
+            title = item.get("title", "No title")
+            snippet = item.get("snippet", item.get("description", ""))
+            formatted += f"  {i}. {title}\n     {snippet[:100]}...\n"
+        return formatted
+
+    def _format_weather_results(self, data: Any) -> str:
+        """Format weather results."""
+        if isinstance(data, dict) and data.get("error"):
+            return f"Weather lookup failed: {data['error']}\n"
+
+        # Assuming weather API returns temp, conditions, etc.
+        temp = data.get("temperature", "unknown")
+        conditions = data.get("conditions", "unknown")
+        location = data.get("location", "requested location")
+
+        return f"Weather for {location}: {temp}, {conditions}\n"
+
+    def _format_codebrain_results(self, data: Any) -> str:
+        """Format codebrain results."""
+        if isinstance(data, dict) and data.get("error"):
+            return f"Codebrain failed: {data['error']}\n"
+
+        # Format code-related results
+        return f"{data}\n"

cortex/context.py

@@ -24,6 +24,7 @@ from neomem_client import NeoMemClient
 # Configuration
 # -----------------------------
 NEOMEM_API = os.getenv("NEOMEM_API", "http://neomem-api:8000")
+NEOMEM_ENABLED = os.getenv("NEOMEM_ENABLED", "false").lower() == "true"
 RELEVANCE_THRESHOLD = float(os.getenv("RELEVANCE_THRESHOLD", "0.4"))
 VERBOSE_DEBUG = os.getenv("VERBOSE_DEBUG", "false").lower() == "true"
 
@@ -148,6 +149,10 @@ async def _search_neomem(
     Returns:
         List of memory objects with full structure, or empty list on failure
     """
+    if not NEOMEM_ENABLED:
+        logger.info("NeoMem search skipped (NEOMEM_ENABLED is false)")
+        return []
+
     try:
         # NeoMemClient reads NEOMEM_API from environment, no base_url parameter
         client = NeoMemClient()
@@ -259,11 +264,15 @@ async def collect_context(session_id: str, user_prompt: str) -> Dict[str, Any]:
         logger.debug(json.dumps(intake_data, indent=2, default=str))
 
     # D. Search NeoMem for relevant memories
-    rag_results = await _search_neomem(
+    if NEOMEM_ENABLED:
-        query=user_prompt,
+        rag_results = await _search_neomem(
-        user_id="brian",  # TODO: Make configurable per session
+            query=user_prompt,
-        limit=5
+            user_id="brian",  # TODO: Make configurable per session
-    )
+            limit=5
+        )
+    else:
+        rag_results = []
+        logger.info("Skipping NeoMem RAG retrieval; NEOMEM_ENABLED is false")
 
     if VERBOSE_DEBUG:
         logger.debug(f"[COLLECT_CONTEXT] NeoMem search returned {len(rag_results)} results")

cortex/data/self_state.json

@@ -0,0 +1,20 @@
+{
+  "mood": "neutral",
+  "energy": 0.8,
+  "focus": "user_request",
+  "confidence": 0.7,
+  "curiosity": 1.0,
+  "last_updated": "2025-12-19T20:25:25.437557",
+  "interaction_count": 16,
+  "learning_queue": [],
+  "active_goals": [],
+  "preferences": {
+    "verbosity": "medium",
+    "formality": "casual",
+    "proactivity": 0.3
+  },
+  "metadata": {
+    "version": "1.0",
+    "created_at": "2025-12-14T03:28:49.364768"
+  }
+}

cortex/intake/intake.py

@@ -234,25 +234,27 @@ def push_to_neomem(summary: str, session_id: str, level: str) -> None:
 async def summarize_context(session_id: str, exchanges: list[dict]):
     """
     Internal summarizer that uses Cortex's LLM router.
-    Produces L1 / L5 / L10 / L20 / L30 summaries.
+    Produces cascading summaries based on exchange count:
+    - L1: Always (most recent activity)
+    - L2: After 2+ exchanges
+    - L5: After 5+ exchanges
+    - L10: After 10+ exchanges
+    - L20: After 20+ exchanges
+    - L30: After 30+ exchanges
 
     Args:
         session_id: The conversation/session ID
         exchanges: A list of {"user_msg": ..., "assistant_msg": ..., "timestamp": ...}
     """
 
-    # Build raw conversation text
+    exchange_count = len(exchanges)
-    convo_lines = []
-    for ex in exchanges:
-        convo_lines.append(f"User: {ex.get('user_msg','')}")
-        convo_lines.append(f"Assistant: {ex.get('assistant_msg','')}")
-    convo_text = "\n".join(convo_lines)
 
-    if not convo_text.strip():
+    if exchange_count == 0:
         return {
             "session_id": session_id,
             "exchange_count": 0,
             "L1": "",
+            "L2": "",
             "L5": "",
             "L10": "",
             "L20": "",
@@ -260,63 +262,54 @@ async def summarize_context(session_id: str, exchanges: list[dict]):
             "last_updated": datetime.now().isoformat()
         }
 
-    # Prompt the LLM (internal — no HTTP)
+    result = {
-    prompt = f"""
+        "session_id": session_id,
-Summarize the conversation below into multiple compression levels.
+        "exchange_count": exchange_count,
-
+        "L1": "",
-Conversation:
+        "L2": "",
-----------------
+        "L5": "",
-{convo_text}
+        "L10": "",
-----------------
+        "L20": "",
-
+        "L30": "",
-Output strictly in JSON with keys:
+        "last_updated": datetime.now().isoformat()
-L1  → ultra short summary (1–2 sentences max)
+    }
-L5  → short summary
-L10 → medium summary
-L20 → detailed overview
-L30 → full detailed summary
-
-JSON only. No text outside JSON.
-"""
 
     try:
-        llm_response = await call_llm(
+        # L1: Always generate (most recent exchanges)
-            prompt,
+        result["L1"] = await summarize_simple(exchanges[-5:])
-            backend=INTAKE_LLM,
+        print(f"[Intake] Generated L1 for {session_id} ({exchange_count} exchanges)")
-            temperature=0.2
-        )
 
-        print(f"[Intake] LLM response length: {len(llm_response) if llm_response else 0}")
+        # L2: After 2+ exchanges
-        print(f"[Intake] LLM response preview: {llm_response[:200] if llm_response else '(empty)'}")
+        if exchange_count >= 2:
+            result["L2"] = await summarize_simple(exchanges[-2:])
+            print(f"[Intake] Generated L2 for {session_id}")
 
-        # LLM should return JSON, parse it
+        # L5: After 5+ exchanges
-        if not llm_response or not llm_response.strip():
+        if exchange_count >= 5:
-            raise ValueError("Empty response from LLM")
+            result["L5"] = await summarize_simple(exchanges[-10:])
+            print(f"[Intake] Generated L5 for {session_id}")
 
-        summary = json.loads(llm_response)
+        # L10: After 10+ exchanges (Reality Check)
+        if exchange_count >= 10:
+            result["L10"] = await summarize_L10(session_id, exchanges)
+            print(f"[Intake] Generated L10 for {session_id}")
 
-        return {
+        # L20: After 20+ exchanges (Session Overview - merges L10s)
-            "session_id": session_id,
+        if exchange_count >= 20 and exchange_count % 10 == 0:
-            "exchange_count": len(exchanges),
+            result["L20"] = await summarize_L20(session_id)
-            "L1": summary.get("L1", ""),
+            print(f"[Intake] Generated L20 for {session_id}")
-            "L5": summary.get("L5", ""),
+
-            "L10": summary.get("L10", ""),
+        # L30: After 30+ exchanges (Continuity Report - merges L20s)
-            "L20": summary.get("L20", ""),
+        if exchange_count >= 30 and exchange_count % 10 == 0:
-            "L30": summary.get("L30", ""),
+            result["L30"] = await summarize_L30(session_id)
-            "last_updated": datetime.now().isoformat()
+            print(f"[Intake] Generated L30 for {session_id}")
-        }
+
+        return result
 
     except Exception as e:
-        return {
+        print(f"[Intake] Error during summarization: {e}")
-            "session_id": session_id,
+        result["L1"] = f"[Error summarizing: {str(e)}]"
-            "exchange_count": len(exchanges),
+        return result
-            "L1": f"[Error summarizing: {str(e)}]",
-            "L5": "",
-            "L10": "",
-            "L20": "",
-            "L30": "",
-            "last_updated": datetime.now().isoformat()
-        }
 
 # ─────────────────────────────────
 # Background summarization stub

cortex/persona/speak.py

@@ -59,17 +59,44 @@ Guidelines:
 # Build persona prompt
 # ============================================================
 
-def build_speak_prompt(final_answer: str) -> str:
+def build_speak_prompt(final_answer: str, tone: str = "neutral", depth: str = "medium") -> str:
     """
     Wrap Cortex's final neutral answer in the Lyra persona.
     Cortex → neutral reasoning
     Speak → stylistic transformation
-    
+
     The LLM sees the original answer and rewrites it in Lyra's voice.
+
+    Args:
+        final_answer: The neutral reasoning output
+        tone: Desired emotional tone (neutral | warm | focused | playful | direct)
+        depth: Response depth (short | medium | deep)
     """
+
+    # Tone-specific guidance
+    tone_guidance = {
+        "neutral": "balanced and professional",
+        "warm": "friendly and empathetic",
+        "focused": "precise and technical",
+        "playful": "light and engaging",
+        "direct": "concise and straightforward"
+    }
+
+    depth_guidance = {
+        "short": "Keep responses brief and to-the-point.",
+        "medium": "Provide balanced detail.",
+        "deep": "Elaborate thoroughly with nuance and examples."
+    }
+
+    tone_hint = tone_guidance.get(tone, "balanced and professional")
+    depth_hint = depth_guidance.get(depth, "Provide balanced detail.")
+
     return f"""
 {PERSONA_STYLE}
 
+Tone guidance: Your response should be {tone_hint}.
+Depth guidance: {depth_hint}
+
 Rewrite the following message into Lyra's natural voice.
 Preserve meaning exactly.
 
@@ -84,16 +111,21 @@ Preserve meaning exactly.
 # Public API — async wrapper
 # ============================================================
 
-async def speak(final_answer: str) -> str:
+async def speak(final_answer: str, tone: str = "neutral", depth: str = "medium") -> str:
     """
     Given the final refined answer from Cortex,
     apply Lyra persona styling using the designated backend.
+
+    Args:
+        final_answer: The polished answer from refinement stage
+        tone: Desired emotional tone (neutral | warm | focused | playful | direct)
+        depth: Response depth (short | medium | deep)
     """
 
     if not final_answer:
         return ""
 
-    prompt = build_speak_prompt(final_answer)
+    prompt = build_speak_prompt(final_answer, tone, depth)
 
     backend = SPEAK_BACKEND
 

cortex/reasoning/reasoning.py

@@ -45,7 +45,9 @@ async def reason_check(
     identity_block: dict | None,
     rag_block: dict | None,
     reflection_notes: list[str],
-    context: dict | None = None
+    context: dict | None = None,
+    monologue: dict | None = None,  # NEW: Inner monologue guidance
+    executive_plan: dict | None = None  # NEW: Executive plan for complex tasks
 ) -> str:
     """
     Build the *draft answer* for Lyra Cortex.
@@ -57,6 +59,8 @@ async def reason_check(
         rag_block: Relevant long-term memories from NeoMem
         reflection_notes: Meta-awareness notes from reflection stage
         context: Unified context state from context.py (session state, intake, rag, etc.)
+        monologue: Inner monologue analysis (intent, tone, depth, consult_executive)
+        executive_plan: Executive plan for complex queries (steps, tools, strategy)
     """
 
     # --------------------------------------------------------
@@ -79,6 +83,52 @@ async def reason_check(
         except Exception:
             identity_txt = f"Identity Rules:\n{str(identity_block)}\n\n"
 
+    # --------------------------------------------------------
+    # Inner Monologue guidance (NEW)
+    # --------------------------------------------------------
+    monologue_section = ""
+    if monologue:
+        intent = monologue.get("intent", "unknown")
+        tone_desired = monologue.get("tone", "neutral")
+        depth_desired = monologue.get("depth", "medium")
+
+        monologue_section = f"""
+=== INNER MONOLOGUE GUIDANCE ===
+User Intent Detected: {intent}
+Desired Tone: {tone_desired}
+Desired Response Depth: {depth_desired}
+
+Adjust your response accordingly:
+- Focus on addressing the {intent} intent
+- Aim for {depth_desired} depth (short/medium/deep)
+- The persona layer will handle {tone_desired} tone, focus on content
+
+"""
+
+    # --------------------------------------------------------
+    # Executive Plan (NEW)
+    # --------------------------------------------------------
+    plan_section = ""
+    if executive_plan:
+        plan_section = f"""
+=== EXECUTIVE PLAN ===
+Task Complexity: {executive_plan.get('estimated_complexity', 'unknown')}
+Plan Summary: {executive_plan.get('summary', 'No summary')}
+
+Detailed Plan:
+{executive_plan.get('plan_text', 'No detailed plan available')}
+
+Required Steps:
+"""
+        for idx, step in enumerate(executive_plan.get('steps', []), 1):
+            plan_section += f"{idx}. {step}\n"
+
+        tools_needed = executive_plan.get('tools_needed', [])
+        if tools_needed:
+            plan_section += f"\nTools to leverage: {', '.join(tools_needed)}\n"
+
+        plan_section += "\nFollow this plan while generating your response.\n\n"
+
     # --------------------------------------------------------
     # RAG block (optional factual grounding)
     # --------------------------------------------------------
@@ -164,6 +214,8 @@ async def reason_check(
     prompt = (
         f"{notes_section}"
         f"{identity_txt}"
+        f"{monologue_section}"  # NEW: Intent/tone/depth guidance
+        f"{plan_section}"  # NEW: Executive plan if generated
         f"{context_txt}"  # Context BEFORE RAG for better coherence
         f"{rag_txt}"
         f"User message:\n{user_prompt}\n\n"

cortex/router.py

@@ -2,7 +2,7 @@
 
 import os
 import logging
-from fastapi import APIRouter, HTTPException
+from fastapi import APIRouter
 from pydantic import BaseModel
 
 from reasoning.reasoning import reason_check
@@ -13,17 +13,19 @@ from persona.identity import load_identity
 from context import collect_context, update_last_assistant_message
 from intake.intake import add_exchange_internal
 
+from autonomy.monologue.monologue import InnerMonologue
+from autonomy.self.state import load_self_state
 
-# -----------------------------
+
-# Debug configuration
+# -------------------------------------------------------------------
-# -----------------------------
+# Setup
+# -------------------------------------------------------------------
 VERBOSE_DEBUG = os.getenv("VERBOSE_DEBUG", "false").lower() == "true"
 logger = logging.getLogger(__name__)
 
 if VERBOSE_DEBUG:
     logger.setLevel(logging.DEBUG)
 
-    # Console handler
     console_handler = logging.StreamHandler()
     console_handler.setFormatter(logging.Formatter(
         '%(asctime)s [ROUTER] %(levelname)s: %(message)s',
@@ -31,7 +33,6 @@ if VERBOSE_DEBUG:
     ))
     logger.addHandler(console_handler)
 
-    # File handler
     try:
         os.makedirs('/app/logs', exist_ok=True)
         file_handler = logging.FileHandler('/app/logs/cortex_verbose_debug.log', mode='a')
@@ -40,28 +41,27 @@ if VERBOSE_DEBUG:
             datefmt='%Y-%m-%d %H:%M:%S'
         ))
         logger.addHandler(file_handler)
-        logger.debug("VERBOSE_DEBUG mode enabled for router.py - logging to file")
+        logger.debug("VERBOSE_DEBUG enabled for router.py")
     except Exception as e:
-        logger.debug(f"VERBOSE_DEBUG mode enabled for router.py - file logging failed: {e}")
+        logger.debug(f"File logging failed: {e}")
+
 
-# -----------------------------
-# Router (NOT FastAPI app)
-# -----------------------------
 cortex_router = APIRouter()
+inner_monologue = InnerMonologue()
 
 
-# -----------------------------
+# -------------------------------------------------------------------
-# Pydantic models
+# Models
-# -----------------------------
+# -------------------------------------------------------------------
 class ReasonRequest(BaseModel):
     session_id: str
     user_prompt: str
     temperature: float | None = None
 
 
-# -----------------------------
+# -------------------------------------------------------------------
 # /reason endpoint
-# -----------------------------
+# -------------------------------------------------------------------
 @cortex_router.post("/reason")
 async def run_reason(req: ReasonRequest):
 
@@ -71,7 +71,9 @@ async def run_reason(req: ReasonRequest):
         logger.debug(f"[PIPELINE START] User prompt: {req.user_prompt[:200]}...")
         logger.debug(f"{'='*80}\n")
 
-    # 0. Collect unified context from all sources
+    # ----------------------------------------------------------------
+    # STAGE 0 — Context
+    # ----------------------------------------------------------------
     if VERBOSE_DEBUG:
         logger.debug("[STAGE 0] Collecting unified context...")
 
@@ -80,7 +82,9 @@ async def run_reason(req: ReasonRequest):
     if VERBOSE_DEBUG:
         logger.debug(f"[STAGE 0] Context collected - {len(context_state.get('rag', []))} RAG results")
 
-    # 0.5. Load identity block
+    # ----------------------------------------------------------------
+    # STAGE 0.5 — Identity
+    # ----------------------------------------------------------------
     if VERBOSE_DEBUG:
         logger.debug("[STAGE 0.5] Loading identity block...")
 
@@ -89,37 +93,133 @@ async def run_reason(req: ReasonRequest):
     if VERBOSE_DEBUG:
         logger.debug(f"[STAGE 0.5] Identity loaded: {identity_block.get('name', 'Unknown')}")
 
-    # 1. Extract Intake summary for reflection
+    # ----------------------------------------------------------------
-    # Use L20 (Session Overview) as primary summary for reflection
+    # STAGE 0.6 — Inner Monologue (observer-only)
+    # ----------------------------------------------------------------
+    if VERBOSE_DEBUG:
+        logger.debug("[STAGE 0.6] Running inner monologue...")
+
+    inner_result = None
+    try:
+        self_state = load_self_state()
+
+        mono_context = {
+            "user_message": req.user_prompt,
+            "session_id": req.session_id,
+            "self_state": self_state,
+            "context_summary": context_state,
+        }
+
+        inner_result = await inner_monologue.process(mono_context)
+        logger.info(f"[INNER_MONOLOGUE] {inner_result}")
+
+        # Store in context for downstream use
+        context_state["monologue"] = inner_result
+
+    except Exception as e:
+        logger.warning(f"[INNER_MONOLOGUE] failed: {e}")
+
+    # ----------------------------------------------------------------
+    # STAGE 0.7 — Executive Planning (conditional)
+    # ----------------------------------------------------------------
+    executive_plan = None
+    if inner_result and inner_result.get("consult_executive"):
+        if VERBOSE_DEBUG:
+            logger.debug("[STAGE 0.7] Executive consultation requested...")
+
+        try:
+            from autonomy.executive.planner import plan_execution
+            executive_plan = await plan_execution(
+                user_prompt=req.user_prompt,
+                intent=inner_result.get("intent", "unknown"),
+                context_state=context_state,
+                identity_block=identity_block
+            )
+            logger.info(f"[EXECUTIVE] Generated plan: {executive_plan.get('summary', 'N/A')}")
+        except Exception as e:
+            logger.warning(f"[EXECUTIVE] Planning failed: {e}")
+            executive_plan = None
+
+    # ----------------------------------------------------------------
+    # STAGE 0.8 — Autonomous Tool Invocation
+    # ----------------------------------------------------------------
+    tool_results = None
+    autonomous_enabled = os.getenv("ENABLE_AUTONOMOUS_TOOLS", "true").lower() == "true"
+    tool_confidence_threshold = float(os.getenv("AUTONOMOUS_TOOL_CONFIDENCE_THRESHOLD", "0.6"))
+
+    if autonomous_enabled and inner_result:
+        if VERBOSE_DEBUG:
+            logger.debug("[STAGE 0.8] Analyzing autonomous tool needs...")
+
+        try:
+            from autonomy.tools.decision_engine import ToolDecisionEngine
+            from autonomy.tools.orchestrator import ToolOrchestrator
+
+            # Analyze which tools to invoke
+            decision_engine = ToolDecisionEngine()
+            tool_decision = await decision_engine.analyze_tool_needs(
+                user_prompt=req.user_prompt,
+                monologue=inner_result,
+                context_state=context_state,
+                available_tools=["RAG", "WEB", "WEATHER", "CODEBRAIN"]
+            )
+
+            # Execute tools if confidence threshold met
+            if tool_decision["should_invoke_tools"] and tool_decision["confidence"] >= tool_confidence_threshold:
+                orchestrator = ToolOrchestrator(tool_timeout=30)
+                tool_results = await orchestrator.execute_tools(
+                    tools_to_invoke=tool_decision["tools_to_invoke"],
+                    context_state=context_state
+                )
+
+                # Format results for context injection
+                tool_context = orchestrator.format_results_for_context(tool_results)
+                context_state["autonomous_tool_results"] = tool_context
+
+                if VERBOSE_DEBUG:
+                    summary = tool_results.get("execution_summary", {})
+                    logger.debug(f"[STAGE 0.8] Tools executed: {summary.get('successful', [])} succeeded")
+            else:
+                if VERBOSE_DEBUG:
+                    logger.debug(f"[STAGE 0.8] No tools invoked (confidence: {tool_decision.get('confidence', 0):.2f})")
+
+        except Exception as e:
+            logger.warning(f"[STAGE 0.8] Autonomous tool invocation failed: {e}")
+            if VERBOSE_DEBUG:
+                import traceback
+                traceback.print_exc()
+
+    # ----------------------------------------------------------------
+    # STAGE 1 — Intake summary
+    # ----------------------------------------------------------------
     intake_summary = "(no context available)"
     if context_state.get("intake"):
-        l20_summary = context_state["intake"].get("L20")
+        l20 = context_state["intake"].get("L20")
-        if l20_summary and isinstance(l20_summary, dict):
+        if isinstance(l20, dict):
-            intake_summary = l20_summary.get("summary", "(no context available)")
+            intake_summary = l20.get("summary", intake_summary)
-        elif isinstance(l20_summary, str):
+        elif isinstance(l20, str):
-            intake_summary = l20_summary
+            intake_summary = l20
 
     if VERBOSE_DEBUG:
         logger.debug(f"[STAGE 1] Intake summary extracted (L20): {intake_summary[:150]}...")
 
-    # 2. Reflection
+    # ----------------------------------------------------------------
+    # STAGE 2 — Reflection
+    # ----------------------------------------------------------------
     if VERBOSE_DEBUG:
         logger.debug("[STAGE 2] Running reflection...")
 
     try:
         reflection = await reflect_notes(intake_summary, identity_block=identity_block)
         reflection_notes = reflection.get("notes", [])
-
-        if VERBOSE_DEBUG:
-            logger.debug(f"[STAGE 2] Reflection complete - {len(reflection_notes)} notes generated")
-            for idx, note in enumerate(reflection_notes, 1):
-                logger.debug(f"  Note {idx}: {note}")
     except Exception as e:
         reflection_notes = []
         if VERBOSE_DEBUG:
             logger.debug(f"[STAGE 2] Reflection failed: {e}")
 
-    # 3. First-pass reasoning draft
+    # ----------------------------------------------------------------
+    # STAGE 3 — Reasoning (draft)
+    # ----------------------------------------------------------------
     if VERBOSE_DEBUG:
         logger.debug("[STAGE 3] Running reasoning (draft)...")
 
@@ -128,14 +228,14 @@ async def run_reason(req: ReasonRequest):
         identity_block=identity_block,
         rag_block=context_state.get("rag", []),
         reflection_notes=reflection_notes,
-        context=context_state
+        context=context_state,
+        monologue=inner_result,  # NEW: Pass monologue guidance
+        executive_plan=executive_plan  # NEW: Pass executive plan
     )
 
-    if VERBOSE_DEBUG:
+    # ----------------------------------------------------------------
-        logger.debug(f"[STAGE 3] Draft answer ({len(draft)} chars):")
+    # STAGE 4 — Refinement
-        logger.debug(f"--- DRAFT START ---\n{draft}\n--- DRAFT END ---")
+    # ----------------------------------------------------------------
-
-    # 4. Refinement
     if VERBOSE_DEBUG:
         logger.debug("[STAGE 4] Running refinement...")
 
@@ -145,35 +245,92 @@ async def run_reason(req: ReasonRequest):
         identity_block=identity_block,
         rag_block=context_state.get("rag", []),
     )
+
     final_neutral = result["final_output"]
 
-    if VERBOSE_DEBUG:
+    # ----------------------------------------------------------------
-        logger.debug(f"[STAGE 4] Refined answer ({len(final_neutral)} chars):")
+    # STAGE 5 — Persona
-        logger.debug(f"--- REFINED START ---\n{final_neutral}\n--- REFINED END ---")
+    # ----------------------------------------------------------------
-
-    # 5. Persona layer
     if VERBOSE_DEBUG:
         logger.debug("[STAGE 5] Applying persona layer...")
 
-    persona_answer = await speak(final_neutral)
+    # Extract tone and depth from monologue for persona guidance
+    tone = inner_result.get("tone", "neutral") if inner_result else "neutral"
+    depth = inner_result.get("depth", "medium") if inner_result else "medium"
 
-    if VERBOSE_DEBUG:
+    persona_answer = await speak(final_neutral, tone=tone, depth=depth)
-        logger.debug(f"[STAGE 5] Persona answer ({len(persona_answer)} chars):")
-        logger.debug(f"--- PERSONA START ---\n{persona_answer}\n--- PERSONA END ---")
-
-    # 6. Update session state with assistant's response
-    if VERBOSE_DEBUG:
-        logger.debug("[STAGE 6] Updating session state...")
 
+    # ----------------------------------------------------------------
+    # STAGE 6 — Session update
+    # ----------------------------------------------------------------
     update_last_assistant_message(req.session_id, persona_answer)
 
+    # ----------------------------------------------------------------
+    # STAGE 6.5 — Self-state update & Pattern Learning
+    # ----------------------------------------------------------------
+    try:
+        from autonomy.self.analyzer import analyze_and_update_state
+        await analyze_and_update_state(
+            monologue=inner_result or {},
+            user_prompt=req.user_prompt,
+            response=persona_answer,
+            context=context_state
+        )
+    except Exception as e:
+        logger.warning(f"[SELF_STATE] Update failed: {e}")
+
+    # Pattern learning
+    try:
+        from autonomy.learning.pattern_learner import get_pattern_learner
+        learner = get_pattern_learner()
+        await learner.learn_from_interaction(
+            user_prompt=req.user_prompt,
+            response=persona_answer,
+            monologue=inner_result or {},
+            context=context_state
+        )
+    except Exception as e:
+        logger.warning(f"[PATTERN_LEARNER] Learning failed: {e}")
+
+    # ----------------------------------------------------------------
+    # STAGE 7 — Proactive Monitoring & Suggestions
+    # ----------------------------------------------------------------
+    proactive_enabled = os.getenv("ENABLE_PROACTIVE_MONITORING", "true").lower() == "true"
+    proactive_min_priority = float(os.getenv("PROACTIVE_SUGGESTION_MIN_PRIORITY", "0.6"))
+
+    if proactive_enabled:
+        try:
+            from autonomy.proactive.monitor import get_proactive_monitor
+
+            monitor = get_proactive_monitor(min_priority=proactive_min_priority)
+            self_state = load_self_state()  # Already imported at top of file
+
+            suggestion = await monitor.analyze_session(
+                session_id=req.session_id,
+                context_state=context_state,
+                self_state=self_state
+            )
+
+            # Append suggestion to response if exists
+            if suggestion:
+                suggestion_text = monitor.format_suggestion(suggestion)
+                persona_answer += suggestion_text
+
+                if VERBOSE_DEBUG:
+                    logger.debug(f"[STAGE 7] Proactive suggestion added: {suggestion['type']} (priority: {suggestion['priority']:.2f})")
+
+        except Exception as e:
+            logger.warning(f"[STAGE 7] Proactive monitoring failed: {e}")
+
     if VERBOSE_DEBUG:
         logger.debug(f"\n{'='*80}")
         logger.debug(f"[PIPELINE COMPLETE] Session: {req.session_id}")
         logger.debug(f"[PIPELINE COMPLETE] Final answer length: {len(persona_answer)} chars")
         logger.debug(f"{'='*80}\n")
 
-    # 7. Return full bundle
+    # ----------------------------------------------------------------
+    # RETURN
+    # ----------------------------------------------------------------
     return {
         "draft": draft,
         "neutral": final_neutral,
@@ -189,9 +346,9 @@ async def run_reason(req: ReasonRequest):
     }
 
 
-# -----------------------------
+# -------------------------------------------------------------------
-# Intake ingest (internal feed)
+# /ingest endpoint (internal)
-# -----------------------------
+# -------------------------------------------------------------------
 class IngestPayload(BaseModel):
     session_id: str
     user_msg: str
@@ -200,107 +357,18 @@ class IngestPayload(BaseModel):
 
 @cortex_router.post("/ingest")
 async def ingest(payload: IngestPayload):
-    """
-    Receives (session_id, user_msg, assistant_msg) from Relay
-    and pushes directly into Intake's in-memory buffer.
-
-    Uses lenient error handling - always returns success to avoid
-    breaking the chat pipeline.
-    """
     try:
-        # 1. Update Cortex session state
         update_last_assistant_message(payload.session_id, payload.assistant_msg)
     except Exception as e:
-        logger.warning(f"[INGEST] Failed to update session state: {e}")
+        logger.warning(f"[INGEST] Session update failed: {e}")
-        # Continue anyway (lenient mode)
 
     try:
-        # 2. Feed Intake internally (no HTTP)
         add_exchange_internal({
             "session_id": payload.session_id,
             "user_msg": payload.user_msg,
             "assistant_msg": payload.assistant_msg,
         })
-        logger.debug(f"[INGEST] Added exchange to Intake for {payload.session_id}")
     except Exception as e:
-        logger.warning(f"[INGEST] Failed to add to Intake: {e}")
+        logger.warning(f"[INGEST] Intake update failed: {e}")
-        # Continue anyway (lenient mode)
-
-    # Always return success (user requirement: never fail chat pipeline)
-    return {
-        "status": "ok",
-        "session_id": payload.session_id
-    }
-
-# -----------------------------
-# Debug endpoint: summarized context
-# -----------------------------
-@cortex_router.get("/debug/summary")
-async def debug_summary(session_id: str):
-    """
-    Diagnostic endpoint that runs Intake's summarize_context() for a session.
-
-    Shows exactly what L1/L5/L10/L20/L30 summaries would look like
-    inside the actual Uvicorn worker, using the real SESSIONS buffer.
-    """
-    from intake.intake import SESSIONS, summarize_context
-
-    # Validate session
-    session = SESSIONS.get(session_id)
-    if not session:
-        return {"error": "session not found", "session_id": session_id}
-
-    # Convert deque into the structure summarize_context expects
-    buffer = session["buffer"]
-    exchanges = [
-        {
-            "user_msg": ex.get("user_msg", ""),
-            "assistant_msg": ex.get("assistant_msg", ""),
-        }
-        for ex in buffer
-    ]
-
-    # 🔥 CRITICAL FIX — summarize_context is async
-    summary = await summarize_context(session_id, exchanges)
-
-    return {
-        "session_id": session_id,
-        "buffer_size": len(buffer),
-        "exchanges_preview": exchanges[-5:],   # last 5 items
-        "summary": summary
-    }
-
-# -----------------------------
-# Debug endpoint for SESSIONS
-# -----------------------------
-@cortex_router.get("/debug/sessions")
-async def debug_sessions():
-    """
-    Diagnostic endpoint to inspect SESSIONS from within the running Uvicorn worker.
-    This shows the actual state of the in-memory SESSIONS dict.
-    """
-    from intake.intake import SESSIONS
-
-    sessions_data = {}
-    for session_id, session_info in SESSIONS.items():
-        buffer = session_info["buffer"]
-        sessions_data[session_id] = {
-            "created_at": session_info["created_at"].isoformat(),
-            "buffer_size": len(buffer),
-            "buffer_maxlen": buffer.maxlen,
-            "recent_exchanges": [
-                {
-                    "user_msg": ex.get("user_msg", "")[:100],
-                    "assistant_msg": ex.get("assistant_msg", "")[:100],
-                    "timestamp": ex.get("timestamp", "")
-                }
-                for ex in list(buffer)[-5:]  # Last 5 exchanges
-            ]
-        }
-
-    return {
-        "sessions_object_id": id(SESSIONS),
-        "total_sessions": len(SESSIONS),
-        "sessions": sessions_data
-    }
 
+    return {"status": "ok", "session_id": payload.session_id}

cortex/tests/__init__.py

@@ -0,0 +1 @@
+"""Tests for Project Lyra Cortex."""

cortex/tests/test_autonomy_phase1.py

@@ -0,0 +1,197 @@
+"""
+Integration tests for Phase 1 autonomy features.
+Tests monologue integration, executive planning, and self-state persistence.
+"""
+
+import asyncio
+import json
+import sys
+import os
+
+# Add parent directory to path for imports
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+
+from autonomy.monologue.monologue import InnerMonologue
+from autonomy.self.state import load_self_state, update_self_state, get_self_state_instance
+from autonomy.executive.planner import plan_execution
+
+
+async def test_monologue_integration():
+    """Test monologue generates valid output."""
+    print("\n" + "="*60)
+    print("TEST 1: Monologue Integration")
+    print("="*60)
+
+    mono = InnerMonologue()
+
+    context = {
+        "user_message": "Explain quantum computing to me like I'm 5",
+        "session_id": "test_001",
+        "self_state": load_self_state(),
+        "context_summary": {"message_count": 5}
+    }
+
+    result = await mono.process(context)
+
+    assert "intent" in result, "Missing intent field"
+    assert "tone" in result, "Missing tone field"
+    assert "depth" in result, "Missing depth field"
+    assert "consult_executive" in result, "Missing consult_executive field"
+
+    print("✓ Monologue integration test passed")
+    print(f"  Result: {json.dumps(result, indent=2)}")
+
+    return result
+
+
+async def test_executive_planning():
+    """Test executive planner generates valid plans."""
+    print("\n" + "="*60)
+    print("TEST 2: Executive Planning")
+    print("="*60)
+
+    plan = await plan_execution(
+        user_prompt="Help me build a distributed system with microservices architecture",
+        intent="technical_implementation",
+        context_state={
+            "tools_available": ["RAG", "WEB", "CODEBRAIN"],
+            "message_count": 3,
+            "minutes_since_last_msg": 2.5,
+            "active_project": None
+        },
+        identity_block={}
+    )
+
+    assert "summary" in plan, "Missing summary field"
+    assert "plan_text" in plan, "Missing plan_text field"
+    assert "steps" in plan, "Missing steps field"
+    assert len(plan["steps"]) > 0, "No steps generated"
+
+    print("✓ Executive planning test passed")
+    print(f"  Plan summary: {plan['summary']}")
+    print(f"  Steps: {len(plan['steps'])}")
+    print(f"  Complexity: {plan.get('estimated_complexity', 'unknown')}")
+
+    return plan
+
+
+def test_self_state_persistence():
+    """Test self-state loads and updates."""
+    print("\n" + "="*60)
+    print("TEST 3: Self-State Persistence")
+    print("="*60)
+
+    state1 = load_self_state()
+    assert "mood" in state1, "Missing mood field"
+    assert "energy" in state1, "Missing energy field"
+    assert "interaction_count" in state1, "Missing interaction_count"
+
+    initial_count = state1.get("interaction_count", 0)
+    print(f"  Initial interaction count: {initial_count}")
+
+    update_self_state(
+        mood_delta=0.1,
+        energy_delta=-0.05,
+        new_focus="testing"
+    )
+
+    state2 = load_self_state()
+    assert state2["interaction_count"] == initial_count + 1, "Interaction count not incremented"
+    assert state2["focus"] == "testing", "Focus not updated"
+
+    print("✓ Self-state persistence test passed")
+    print(f"  New interaction count: {state2['interaction_count']}")
+    print(f"  New focus: {state2['focus']}")
+    print(f"  New energy: {state2['energy']:.2f}")
+
+    return state2
+
+
+async def test_end_to_end_flow():
+    """Test complete flow from monologue through planning."""
+    print("\n" + "="*60)
+    print("TEST 4: End-to-End Flow")
+    print("="*60)
+
+    # Step 1: Monologue detects complex query
+    mono = InnerMonologue()
+    mono_result = await mono.process({
+        "user_message": "Design a scalable ML pipeline with CI/CD integration",
+        "session_id": "test_e2e",
+        "self_state": load_self_state(),
+        "context_summary": {}
+    })
+
+    print(f"  Monologue intent: {mono_result.get('intent')}")
+    print(f"  Consult executive: {mono_result.get('consult_executive')}")
+
+    # Step 2: If executive requested, generate plan
+    if mono_result.get("consult_executive"):
+        plan = await plan_execution(
+            user_prompt="Design a scalable ML pipeline with CI/CD integration",
+            intent=mono_result.get("intent", "unknown"),
+            context_state={"tools_available": ["CODEBRAIN", "WEB"]},
+            identity_block={}
+        )
+
+        assert plan is not None, "Plan should be generated"
+        print(f"  Executive plan generated: {len(plan.get('steps', []))} steps")
+
+    # Step 3: Update self-state
+    update_self_state(
+        energy_delta=-0.1,  # Complex task is tiring
+        new_focus="ml_pipeline_design",
+        confidence_delta=0.05
+    )
+
+    state = load_self_state()
+    assert state["focus"] == "ml_pipeline_design", "Focus should be updated"
+
+    print("✓ End-to-end flow test passed")
+    print(f"  Final state: {state['mood']}, energy={state['energy']:.2f}")
+
+    return True
+
+
+async def run_all_tests():
+    """Run all Phase 1 tests."""
+    print("\n" + "="*60)
+    print("PHASE 1 AUTONOMY TESTS")
+    print("="*60)
+
+    try:
+        # Test 1: Monologue
+        mono_result = await test_monologue_integration()
+
+        # Test 2: Executive Planning
+        plan_result = await test_executive_planning()
+
+        # Test 3: Self-State
+        state_result = test_self_state_persistence()
+
+        # Test 4: End-to-End
+        await test_end_to_end_flow()
+
+        print("\n" + "="*60)
+        print("ALL TESTS PASSED ✓")
+        print("="*60)
+
+        print("\nSummary:")
+        print(f"  - Monologue: {mono_result.get('intent')} ({mono_result.get('tone')})")
+        print(f"  - Executive: {plan_result.get('estimated_complexity')} complexity")
+        print(f"  - Self-state: {state_result.get('interaction_count')} interactions")
+
+        return True
+
+    except Exception as e:
+        print("\n" + "="*60)
+        print(f"TEST FAILED: {e}")
+        print("="*60)
+        import traceback
+        traceback.print_exc()
+        return False
+
+
+if __name__ == "__main__":
+    success = asyncio.run(run_all_tests())
+    sys.exit(0 if success else 1)

cortex/tests/test_autonomy_phase2.py

@@ -0,0 +1,495 @@
+"""
+Integration tests for Phase 2 autonomy features.
+Tests autonomous tool invocation, proactive monitoring, actions, and pattern learning.
+"""
+
+import asyncio
+import json
+import sys
+import os
+
+# Add parent directory to path for imports
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+
+# Override self-state file path for testing
+os.environ["SELF_STATE_FILE"] = "/tmp/test_self_state.json"
+
+from autonomy.tools.decision_engine import ToolDecisionEngine
+from autonomy.tools.orchestrator import ToolOrchestrator
+from autonomy.proactive.monitor import ProactiveMonitor
+from autonomy.actions.autonomous_actions import AutonomousActionManager
+from autonomy.learning.pattern_learner import PatternLearner
+from autonomy.self.state import load_self_state, get_self_state_instance
+
+
+async def test_tool_decision_engine():
+    """Test autonomous tool decision making."""
+    print("\n" + "="*60)
+    print("TEST 1: Tool Decision Engine")
+    print("="*60)
+
+    engine = ToolDecisionEngine()
+
+    # Test 1a: Memory reference detection
+    result = await engine.analyze_tool_needs(
+        user_prompt="What did we discuss earlier about Python?",
+        monologue={"intent": "clarification", "consult_executive": False},
+        context_state={},
+        available_tools=["RAG", "WEB", "WEATHER"]
+    )
+
+    assert result["should_invoke_tools"], "Should invoke tools for memory reference"
+    assert any(t["tool"] == "RAG" for t in result["tools_to_invoke"]), "Should recommend RAG"
+    assert result["confidence"] > 0.8, f"Confidence should be high for clear memory reference: {result['confidence']}"
+
+    print(f"  ✓ Memory reference detection passed")
+    print(f"    Tools: {[t['tool'] for t in result['tools_to_invoke']]}")
+    print(f"    Confidence: {result['confidence']:.2f}")
+
+    # Test 1b: Web search detection
+    result = await engine.analyze_tool_needs(
+        user_prompt="What's the latest news about AI developments?",
+        monologue={"intent": "information_seeking", "consult_executive": False},
+        context_state={},
+        available_tools=["RAG", "WEB", "WEATHER"]
+    )
+
+    assert result["should_invoke_tools"], "Should invoke tools for current info request"
+    assert any(t["tool"] == "WEB" for t in result["tools_to_invoke"]), "Should recommend WEB"
+
+    print(f"  ✓ Web search detection passed")
+    print(f"    Tools: {[t['tool'] for t in result['tools_to_invoke']]}")
+
+    # Test 1c: Weather detection
+    result = await engine.analyze_tool_needs(
+        user_prompt="What's the weather like today in Boston?",
+        monologue={"intent": "information_seeking", "consult_executive": False},
+        context_state={},
+        available_tools=["RAG", "WEB", "WEATHER"]
+    )
+
+    assert result["should_invoke_tools"], "Should invoke tools for weather query"
+    assert any(t["tool"] == "WEATHER" for t in result["tools_to_invoke"]), "Should recommend WEATHER"
+
+    print(f"  ✓ Weather detection passed")
+
+    # Test 1d: Proactive RAG for complex queries
+    result = await engine.analyze_tool_needs(
+        user_prompt="Design a microservices architecture",
+        monologue={"intent": "technical_implementation", "consult_executive": True},
+        context_state={},
+        available_tools=["RAG", "WEB", "CODEBRAIN"]
+    )
+
+    assert result["should_invoke_tools"], "Should proactively invoke tools for complex queries"
+    rag_tools = [t for t in result["tools_to_invoke"] if t["tool"] == "RAG"]
+    assert len(rag_tools) > 0, "Should include proactive RAG"
+
+    print(f"  ✓ Proactive RAG detection passed")
+    print(f"    Reason: {rag_tools[0]['reason']}")
+
+    print("\n✓ Tool Decision Engine tests passed\n")
+    return result
+
+
+async def test_tool_orchestrator():
+    """Test tool orchestration (mock mode)."""
+    print("\n" + "="*60)
+    print("TEST 2: Tool Orchestrator (Mock Mode)")
+    print("="*60)
+
+    orchestrator = ToolOrchestrator(tool_timeout=5)
+
+    # Since actual tools may not be available, test the orchestrator structure
+    print(f"  Available tools: {list(orchestrator.available_tools.keys())}")
+
+    # Test with tools_to_invoke (will fail gracefully if tools unavailable)
+    tools_to_invoke = [
+        {"tool": "RAG", "query": "test query", "reason": "testing", "priority": 0.9}
+    ]
+
+    result = await orchestrator.execute_tools(
+        tools_to_invoke=tools_to_invoke,
+        context_state={"session_id": "test"}
+    )
+
+    assert "results" in result, "Should return results dict"
+    assert "execution_summary" in result, "Should return execution summary"
+
+    summary = result["execution_summary"]
+    assert "tools_invoked" in summary, "Summary should include tools_invoked"
+    assert "total_time_ms" in summary, "Summary should include timing"
+
+    print(f"  ✓ Orchestrator structure valid")
+    print(f"    Summary: {summary}")
+
+    # Test result formatting
+    formatted = orchestrator.format_results_for_context(result)
+    assert isinstance(formatted, str), "Should format results as string"
+
+    print(f"  ✓ Result formatting works")
+    print(f"    Formatted length: {len(formatted)} chars")
+
+    print("\n✓ Tool Orchestrator tests passed\n")
+    return result
+
+
+async def test_proactive_monitor():
+    """Test proactive monitoring and suggestions."""
+    print("\n" + "="*60)
+    print("TEST 3: Proactive Monitor")
+    print("="*60)
+
+    monitor = ProactiveMonitor(min_priority=0.6)
+
+    # Test 3a: Long silence detection
+    context_state = {
+        "message_count": 5,
+        "minutes_since_last_msg": 35  # > 30 minutes
+    }
+
+    self_state = load_self_state()
+
+    suggestion = await monitor.analyze_session(
+        session_id="test_silence",
+        context_state=context_state,
+        self_state=self_state
+    )
+
+    assert suggestion is not None, "Should generate suggestion for long silence"
+    assert suggestion["type"] == "check_in", f"Should be check_in type: {suggestion['type']}"
+    assert suggestion["priority"] >= 0.6, "Priority should meet threshold"
+
+    print(f"  ✓ Long silence detection passed")
+    print(f"    Type: {suggestion['type']}, Priority: {suggestion['priority']:.2f}")
+    print(f"    Suggestion: {suggestion['suggestion'][:50]}...")
+
+    # Test 3b: Learning opportunity (high curiosity)
+    self_state["curiosity"] = 0.8
+    self_state["learning_queue"] = ["quantum computing", "rust programming"]
+
+    # Reset cooldown for this test
+    monitor.reset_cooldown("test_learning")
+
+    suggestion = await monitor.analyze_session(
+        session_id="test_learning",
+        context_state={"message_count": 3, "minutes_since_last_msg": 2},
+        self_state=self_state
+    )
+
+    assert suggestion is not None, "Should generate learning suggestion"
+    assert suggestion["type"] == "learning", f"Should be learning type: {suggestion['type']}"
+
+    print(f"  ✓ Learning opportunity detection passed")
+    print(f"    Suggestion: {suggestion['suggestion'][:70]}...")
+
+    # Test 3c: Conversation milestone
+    monitor.reset_cooldown("test_milestone")
+
+    # Reset curiosity to avoid learning suggestion taking precedence
+    self_state["curiosity"] = 0.5
+    self_state["learning_queue"] = []
+
+    suggestion = await monitor.analyze_session(
+        session_id="test_milestone",
+        context_state={"message_count": 50, "minutes_since_last_msg": 1},
+        self_state=self_state
+    )
+
+    assert suggestion is not None, "Should generate milestone suggestion"
+    # Note: learning or summary both valid - check it's a reasonable suggestion
+    assert suggestion["type"] in ["summary", "learning", "check_in"], f"Should be valid type: {suggestion['type']}"
+
+    print(f"  ✓ Conversation milestone detection passed (type: {suggestion['type']})")
+
+    # Test 3d: Cooldown mechanism
+    # Try to get another suggestion immediately (should be blocked)
+    suggestion2 = await monitor.analyze_session(
+        session_id="test_milestone",
+        context_state={"message_count": 51, "minutes_since_last_msg": 1},
+        self_state=self_state
+    )
+
+    assert suggestion2 is None, "Should not generate suggestion during cooldown"
+
+    print(f"  ✓ Cooldown mechanism working")
+
+    # Check stats
+    stats = monitor.get_session_stats("test_milestone")
+    assert stats["cooldown_active"], "Cooldown should be active"
+    print(f"    Cooldown remaining: {stats['cooldown_remaining']}s")
+
+    print("\n✓ Proactive Monitor tests passed\n")
+    return suggestion
+
+
+async def test_autonomous_actions():
+    """Test autonomous action execution."""
+    print("\n" + "="*60)
+    print("TEST 4: Autonomous Actions")
+    print("="*60)
+
+    manager = AutonomousActionManager()
+
+    # Test 4a: List allowed actions
+    allowed = manager.get_allowed_actions()
+    assert "create_memory" in allowed, "Should have create_memory action"
+    assert "update_goal" in allowed, "Should have update_goal action"
+    assert "learn_topic" in allowed, "Should have learn_topic action"
+
+    print(f"  ✓ Allowed actions: {allowed}")
+
+    # Test 4b: Validate actions
+    validation = manager.validate_action("create_memory", {"text": "test memory"})
+    assert validation["valid"], "Should validate correct action"
+
+    print(f"  ✓ Action validation passed")
+
+    # Test 4c: Execute learn_topic action
+    result = await manager.execute_action(
+        action_type="learn_topic",
+        parameters={"topic": "rust programming", "reason": "testing", "priority": 0.8},
+        context={"session_id": "test"}
+    )
+
+    assert result["success"], f"Action should succeed: {result.get('error', 'unknown')}"
+    assert "topic" in result["result"], "Should return topic info"
+
+    print(f"  ✓ learn_topic action executed")
+    print(f"    Topic: {result['result']['topic']}")
+    print(f"    Queue position: {result['result']['queue_position']}")
+
+    # Test 4d: Execute update_focus action
+    result = await manager.execute_action(
+        action_type="update_focus",
+        parameters={"focus": "autonomy_testing", "reason": "running tests"},
+        context={"session_id": "test"}
+    )
+
+    assert result["success"], "update_focus should succeed"
+
+    print(f"  ✓ update_focus action executed")
+    print(f"    New focus: {result['result']['new_focus']}")
+
+    # Test 4e: Reject non-whitelisted action
+    result = await manager.execute_action(
+        action_type="delete_all_files",  # NOT in whitelist
+        parameters={},
+        context={"session_id": "test"}
+    )
+
+    assert not result["success"], "Should reject non-whitelisted action"
+    assert "not in whitelist" in result["error"], "Should indicate whitelist violation"
+
+    print(f"  ✓ Non-whitelisted action rejected")
+
+    # Test 4f: Action log
+    log = manager.get_action_log(limit=10)
+    assert len(log) >= 2, f"Should have logged multiple actions (got {len(log)})"
+
+    print(f"  ✓ Action log contains {len(log)} entries")
+
+    print("\n✓ Autonomous Actions tests passed\n")
+    return result
+
+
+async def test_pattern_learner():
+    """Test pattern learning system."""
+    print("\n" + "="*60)
+    print("TEST 5: Pattern Learner")
+    print("="*60)
+
+    # Use temp file for testing
+    test_file = "/tmp/test_patterns.json"
+    learner = PatternLearner(patterns_file=test_file)
+
+    # Test 5a: Learn from multiple interactions
+    for i in range(5):
+        await learner.learn_from_interaction(
+            user_prompt=f"Help me with Python coding task {i}",
+            response=f"Here's help with task {i}...",
+            monologue={"intent": "coding_help", "tone": "focused", "depth": "medium"},
+            context={"session_id": "test", "executive_plan": None}
+        )
+
+    print(f"  ✓ Learned from 5 interactions")
+
+    # Test 5b: Get top topics
+    top_topics = learner.get_top_topics(limit=5)
+    assert len(top_topics) > 0, "Should have learned topics"
+    assert "coding_help" == top_topics[0][0], "coding_help should be top topic"
+
+    print(f"  ✓ Top topics: {[t[0] for t in top_topics[:3]]}")
+
+    # Test 5c: Get preferred tone
+    preferred_tone = learner.get_preferred_tone()
+    assert preferred_tone == "focused", "Should detect focused as preferred tone"
+
+    print(f"  ✓ Preferred tone: {preferred_tone}")
+
+    # Test 5d: Get preferred depth
+    preferred_depth = learner.get_preferred_depth()
+    assert preferred_depth == "medium", "Should detect medium as preferred depth"
+
+    print(f"  ✓ Preferred depth: {preferred_depth}")
+
+    # Test 5e: Get insights
+    insights = learner.get_insights()
+    assert insights["total_interactions"] == 5, "Should track interaction count"
+    assert insights["preferred_tone"] == "focused", "Insights should include tone"
+
+    print(f"  ✓ Insights generated:")
+    print(f"    Total interactions: {insights['total_interactions']}")
+    print(f"    Recommendations: {insights['learning_recommendations']}")
+
+    # Test 5f: Export patterns
+    exported = learner.export_patterns()
+    assert "topic_frequencies" in exported, "Should export all patterns"
+
+    print(f"  ✓ Patterns exported ({len(exported)} keys)")
+
+    # Cleanup
+    if os.path.exists(test_file):
+        os.remove(test_file)
+
+    print("\n✓ Pattern Learner tests passed\n")
+    return insights
+
+
+async def test_end_to_end_autonomy():
+    """Test complete autonomous flow."""
+    print("\n" + "="*60)
+    print("TEST 6: End-to-End Autonomy Flow")
+    print("="*60)
+
+    # Simulate a complex user query that triggers multiple autonomous systems
+    user_prompt = "Remember what we discussed about machine learning? I need current research on transformers."
+
+    monologue = {
+        "intent": "technical_research",
+        "tone": "focused",
+        "depth": "deep",
+        "consult_executive": True
+    }
+
+    context_state = {
+        "session_id": "e2e_test",
+        "message_count": 15,
+        "minutes_since_last_msg": 5
+    }
+
+    print(f"  User prompt: {user_prompt}")
+    print(f"  Monologue intent: {monologue['intent']}")
+
+    # Step 1: Tool decision engine
+    engine = ToolDecisionEngine()
+    tool_decision = await engine.analyze_tool_needs(
+        user_prompt=user_prompt,
+        monologue=monologue,
+        context_state=context_state,
+        available_tools=["RAG", "WEB", "CODEBRAIN"]
+    )
+
+    print(f"\n  Step 1: Tool Decision")
+    print(f"    Should invoke: {tool_decision['should_invoke_tools']}")
+    print(f"    Tools: {[t['tool'] for t in tool_decision['tools_to_invoke']]}")
+    assert tool_decision["should_invoke_tools"], "Should invoke tools"
+    assert len(tool_decision["tools_to_invoke"]) >= 2, "Should recommend multiple tools (RAG + WEB)"
+
+    # Step 2: Pattern learning
+    learner = PatternLearner(patterns_file="/tmp/e2e_test_patterns.json")
+    await learner.learn_from_interaction(
+        user_prompt=user_prompt,
+        response="Here's information about transformers...",
+        monologue=monologue,
+        context=context_state
+    )
+
+    print(f"\n  Step 2: Pattern Learning")
+    top_topics = learner.get_top_topics(limit=3)
+    print(f"    Learned topics: {[t[0] for t in top_topics]}")
+
+    # Step 3: Autonomous action
+    action_manager = AutonomousActionManager()
+    action_result = await action_manager.execute_action(
+        action_type="learn_topic",
+        parameters={"topic": "transformer architectures", "reason": "user interest detected"},
+        context=context_state
+    )
+
+    print(f"\n  Step 3: Autonomous Action")
+    print(f"    Action: learn_topic")
+    print(f"    Success: {action_result['success']}")
+
+    # Step 4: Proactive monitoring (won't trigger due to low message count)
+    monitor = ProactiveMonitor(min_priority=0.6)
+    monitor.reset_cooldown("e2e_test")
+
+    suggestion = await monitor.analyze_session(
+        session_id="e2e_test",
+        context_state=context_state,
+        self_state=load_self_state()
+    )
+
+    print(f"\n  Step 4: Proactive Monitoring")
+    print(f"    Suggestion: {suggestion['type'] if suggestion else 'None (expected for low message count)'}")
+
+    # Cleanup
+    if os.path.exists("/tmp/e2e_test_patterns.json"):
+        os.remove("/tmp/e2e_test_patterns.json")
+
+    print("\n✓ End-to-End Autonomy Flow tests passed\n")
+    return True
+
+
+async def run_all_tests():
+    """Run all Phase 2 tests."""
+    print("\n" + "="*60)
+    print("PHASE 2 AUTONOMY TESTS")
+    print("="*60)
+
+    try:
+        # Test 1: Tool Decision Engine
+        await test_tool_decision_engine()
+
+        # Test 2: Tool Orchestrator
+        await test_tool_orchestrator()
+
+        # Test 3: Proactive Monitor
+        await test_proactive_monitor()
+
+        # Test 4: Autonomous Actions
+        await test_autonomous_actions()
+
+        # Test 5: Pattern Learner
+        await test_pattern_learner()
+
+        # Test 6: End-to-End
+        await test_end_to_end_autonomy()
+
+        print("\n" + "="*60)
+        print("ALL PHASE 2 TESTS PASSED ✓")
+        print("="*60)
+
+        print("\nPhase 2 Features Validated:")
+        print("  ✓ Autonomous tool decision making")
+        print("  ✓ Tool orchestration and execution")
+        print("  ✓ Proactive monitoring and suggestions")
+        print("  ✓ Safe autonomous actions")
+        print("  ✓ Pattern learning and adaptation")
+        print("  ✓ End-to-end autonomous flow")
+
+        return True
+
+    except Exception as e:
+        print("\n" + "="*60)
+        print(f"TEST FAILED: {e}")
+        print("="*60)
+        import traceback
+        traceback.print_exc()
+        return False
+
+
+if __name__ == "__main__":
+    success = asyncio.run(run_all_tests())
+    sys.exit(0 if success else 1)

neomem/neomem/vector_stores/qdrant.py

@@ -1,7 +1,9 @@
 import logging
 import os
 import shutil
+from typing import Optional
 
+from pydantic import BaseModel
 from qdrant_client import QdrantClient
 from qdrant_client.models import (
     Distance,
@@ -19,6 +21,13 @@ from mem0.vector_stores.base import VectorStoreBase
 logger = logging.getLogger(__name__)
 
 
+class OutputData(BaseModel):
+    """Standard output format for vector search results."""
+    id: Optional[str]
+    score: Optional[float]
+    payload: Optional[dict]
+
+
 class Qdrant(VectorStoreBase):
     def __init__(
         self,
@@ -170,7 +179,7 @@ class Qdrant(VectorStoreBase):
             filters (dict, optional): Filters to apply to the search. Defaults to None.
 
         Returns:
-            list: Search results.
+            list: Search results wrapped in OutputData format.
         """
         query_filter = self._create_filter(filters) if filters else None
         hits = self.client.query_points(
@@ -179,7 +188,16 @@ class Qdrant(VectorStoreBase):
             query_filter=query_filter,
             limit=limit,
         )
-        return hits.points
+
+        # Wrap results in OutputData format to match other vector stores
+        return [
+            OutputData(
+                id=str(hit.id),
+                score=hit.score,
+                payload=hit.payload
+            )
+            for hit in hits.points
+        ]
 
     def delete(self, vector_id: int):
         """
@@ -207,7 +225,7 @@ class Qdrant(VectorStoreBase):
         point = PointStruct(id=vector_id, vector=vector, payload=payload)
         self.client.upsert(collection_name=self.collection_name, points=[point])
 
-    def get(self, vector_id: int) -> dict:
+    def get(self, vector_id: int) -> OutputData:
         """
         Retrieve a vector by ID.
 
@@ -215,10 +233,17 @@ class Qdrant(VectorStoreBase):
             vector_id (int): ID of the vector to retrieve.
 
         Returns:
-            dict: Retrieved vector.
+            OutputData: Retrieved vector wrapped in OutputData format.
         """
         result = self.client.retrieve(collection_name=self.collection_name, ids=[vector_id], with_payload=True)
-        return result[0] if result else None
+        if result:
+            hit = result[0]
+            return OutputData(
+                id=str(hit.id),
+                score=None,  # No score for direct retrieval
+                payload=hit.payload
+            )
+        return None
 
     def list_cols(self) -> list:
         """
@@ -251,7 +276,7 @@ class Qdrant(VectorStoreBase):
             limit (int, optional): Number of vectors to return. Defaults to 100.
 
         Returns:
-            list: List of vectors.
+            list: List of vectors wrapped in OutputData format.
         """
         query_filter = self._create_filter(filters) if filters else None
         result = self.client.scroll(
@@ -261,7 +286,18 @@ class Qdrant(VectorStoreBase):
             with_payload=True,
             with_vectors=False,
         )
-        return result
+
+        # Wrap results in OutputData format
+        # scroll() returns tuple: (points, next_page_offset)
+        points = result[0] if isinstance(result, tuple) else result
+        return [
+            OutputData(
+                id=str(point.id),
+                score=None,  # No score for list operation
+                payload=point.payload
+            )
+            for point in points
+        ]
 
     def reset(self):
         """Reset the index by deleting and recreating it."""