omni-cortex 1.17.2__py3-none-any.whl → 1.17.4__py3-none-any.whl

omni_cortex/_bundled/dashboard/backend/database.py

@@ -0,0 +1,1773 @@
+"""Database query functions for reading omni-cortex SQLite databases."""
+
+import json
+import sqlite3
+from collections import Counter
+from datetime import datetime, timedelta
+from pathlib import Path
+from typing import Optional
+
+from models import Activity, FilterParams, Memory, MemoryStats, MemoryUpdate, Session, TimelineEntry
+
+
+def get_connection(db_path: str) -> sqlite3.Connection:
+    """Get a read-only connection to the database."""
+    conn = sqlite3.connect(f"file:{db_path}?mode=ro", uri=True)
+    conn.row_factory = sqlite3.Row
+    return conn
+
+
+def get_write_connection(db_path: str) -> sqlite3.Connection:
+    """Get a writable connection to the database."""
+    conn = sqlite3.connect(db_path)
+    conn.row_factory = sqlite3.Row
+    return conn
+
+
+def ensure_migrations(db_path: str) -> None:
+    """Ensure database has latest migrations applied.
+
+    This function checks for and applies any missing schema updates,
+    including command analytics columns and natural language summary columns.
+    """
+    conn = get_write_connection(db_path)
+
+    # Check if activities table exists
+    table_check = conn.execute(
+        "SELECT name FROM sqlite_master WHERE type='table' AND name='activities'"
+    ).fetchone()
+
+    if not table_check:
+        conn.close()
+        return
+
+    # Check available columns
+    columns = conn.execute("PRAGMA table_info(activities)").fetchall()
+    column_names = {col[1] for col in columns}
+
+    migrations_applied = []
+
+    # Migration v1.1: Command analytics columns
+    if "command_name" not in column_names:
+        conn.executescript("""
+            ALTER TABLE activities ADD COLUMN command_name TEXT;
+            ALTER TABLE activities ADD COLUMN command_scope TEXT;
+            ALTER TABLE activities ADD COLUMN mcp_server TEXT;
+            ALTER TABLE activities ADD COLUMN skill_name TEXT;
+
+            CREATE INDEX IF NOT EXISTS idx_activities_command ON activities(command_name);
+            CREATE INDEX IF NOT EXISTS idx_activities_mcp ON activities(mcp_server);
+            CREATE INDEX IF NOT EXISTS idx_activities_skill ON activities(skill_name);
+        """)
+        migrations_applied.append("v1.1: command analytics columns")
+
+    # Migration v1.2: Natural language summary columns
+    if "summary" not in column_names:
+        conn.executescript("""
+            ALTER TABLE activities ADD COLUMN summary TEXT;
+            ALTER TABLE activities ADD COLUMN summary_detail TEXT;
+        """)
+        migrations_applied.append("v1.2: summary columns")
+
+    if migrations_applied:
+        conn.commit()
+        print(f"[Database] Applied migrations: {', '.join(migrations_applied)}")
+
+    conn.close()
+
+
+def parse_tags(tags_str: Optional[str]) -> list[str]:
+    """Parse tags from JSON string."""
+    if not tags_str:
+        return []
+    try:
+        tags = json.loads(tags_str)
+        return tags if isinstance(tags, list) else []
+    except (json.JSONDecodeError, TypeError):
+        return []
+
+
+def get_memories(db_path: str, filters: FilterParams) -> list[Memory]:
+    """Get memories with filtering, sorting, and pagination."""
+    conn = get_connection(db_path)
+
+    # Build query
+    query = "SELECT * FROM memories WHERE 1=1"
+    params: list = []
+
+    if filters.memory_type:
+        query += " AND type = ?"
+        params.append(filters.memory_type)
+
+    if filters.status:
+        query += " AND status = ?"
+        params.append(filters.status)
+
+    if filters.min_importance is not None:
+        query += " AND importance_score >= ?"
+        params.append(filters.min_importance)
+
+    if filters.max_importance is not None:
+        query += " AND importance_score <= ?"
+        params.append(filters.max_importance)
+
+    if filters.search:
+        query += " AND (content LIKE ? OR context LIKE ?)"
+        search_term = f"%{filters.search}%"
+        params.extend([search_term, search_term])
+
+    # Sorting
+    valid_sort_columns = ["created_at", "last_accessed", "importance_score", "access_count"]
+    sort_by = filters.sort_by if filters.sort_by in valid_sort_columns else "last_accessed"
+    sort_order = "DESC" if filters.sort_order.lower() == "desc" else "ASC"
+    query += f" ORDER BY {sort_by} {sort_order}"
+
+    # Pagination
+    query += " LIMIT ? OFFSET ?"
+    params.extend([filters.limit, filters.offset])
+
+    cursor = conn.execute(query, params)
+    rows = cursor.fetchall()
+
+    memories = []
+    for row in rows:
+        # Parse tags from JSON string
+        tags = parse_tags(row["tags"])
+
+        memories.append(
+            Memory(
+                id=row["id"],
+                content=row["content"],
+                context=row["context"],
+                type=row["type"],
+                status=row["status"] or "fresh",
+                importance_score=int(row["importance_score"] or 50),
+                access_count=row["access_count"] or 0,
+                created_at=datetime.fromisoformat(row["created_at"]),
+                last_accessed=datetime.fromisoformat(row["last_accessed"]) if row["last_accessed"] else None,
+                tags=tags,
+            )
+        )
+
+    conn.close()
+    return memories
+
+
+def get_memory_by_id(db_path: str, memory_id: str) -> Optional[Memory]:
+    """Get a single memory by ID."""
+    conn = get_connection(db_path)
+
+    cursor = conn.execute("SELECT * FROM memories WHERE id = ?", (memory_id,))
+    row = cursor.fetchone()
+
+    if not row:
+        conn.close()
+        return None
+
+    # Parse tags from JSON string
+    tags = parse_tags(row["tags"])
+
+    memory = Memory(
+        id=row["id"],
+        content=row["content"],
+        context=row["context"],
+        type=row["type"],
+        status=row["status"] or "fresh",
+        importance_score=int(row["importance_score"] or 50),
+        access_count=row["access_count"] or 0,
+        created_at=datetime.fromisoformat(row["created_at"]),
+        last_accessed=datetime.fromisoformat(row["last_accessed"]) if row["last_accessed"] else None,
+        tags=tags,
+    )
+
+    conn.close()
+    return memory
+
+
+def get_memory_stats(db_path: str) -> MemoryStats:
+    """Get statistics about memories in the database."""
+    conn = get_connection(db_path)
+
+    # Total count
+    total = conn.execute("SELECT COUNT(*) FROM memories").fetchone()[0]
+
+    # By type
+    type_cursor = conn.execute("SELECT type, COUNT(*) as count FROM memories GROUP BY type")
+    by_type = {row["type"]: row["count"] for row in type_cursor.fetchall()}
+
+    # By status
+    status_cursor = conn.execute("SELECT status, COUNT(*) as count FROM memories GROUP BY status")
+    by_status = {(row["status"] or "fresh"): row["count"] for row in status_cursor.fetchall()}
+
+    # Average importance
+    avg_cursor = conn.execute("SELECT AVG(importance_score) FROM memories")
+    avg_importance = avg_cursor.fetchone()[0] or 0.0
+
+    # Total access count
+    access_cursor = conn.execute("SELECT SUM(access_count) FROM memories")
+    total_access = access_cursor.fetchone()[0] or 0
+
+    # Tags with counts - extract from JSON column
+    tags_cursor = conn.execute("SELECT tags FROM memories WHERE tags IS NOT NULL AND tags != ''")
+    tag_counter: Counter = Counter()
+    for row in tags_cursor.fetchall():
+        tags = parse_tags(row["tags"])
+        tag_counter.update(tags)
+
+    tags = [{"name": name, "count": count} for name, count in tag_counter.most_common(50)]
+
+    conn.close()
+
+    return MemoryStats(
+        total_count=total,
+        by_type=by_type,
+        by_status=by_status,
+        avg_importance=round(avg_importance, 1),
+        total_access_count=total_access,
+        tags=tags,
+    )
+
+
+def get_activities(
+    db_path: str,
+    event_type: Optional[str] = None,
+    tool_name: Optional[str] = None,
+    limit: int = 100,
+    offset: int = 0,
+) -> list[Activity]:
+    """Get activity log entries with all available fields."""
+    conn = get_connection(db_path)
+
+    # Check available columns for backward compatibility
+    columns = conn.execute("PRAGMA table_info(activities)").fetchall()
+    column_names = {col[1] for col in columns}
+
+    query = "SELECT * FROM activities WHERE 1=1"
+    params: list = []
+
+    if event_type:
+        query += " AND event_type = ?"
+        params.append(event_type)
+
+    if tool_name:
+        query += " AND tool_name = ?"
+        params.append(tool_name)
+
+    query += " ORDER BY timestamp DESC LIMIT ? OFFSET ?"
+    params.extend([limit, offset])
+
+    cursor = conn.execute(query, params)
+    activities = []
+
+    for row in cursor.fetchall():
+        # Parse timestamp - handle both with and without timezone
+        ts_str = row["timestamp"]
+        try:
+            ts = datetime.fromisoformat(ts_str)
+        except ValueError:
+            # Fallback for edge cases
+            ts = datetime.now()
+
+        activity_data = {
+            "id": row["id"],
+            "session_id": row["session_id"],
+            "event_type": row["event_type"],
+            "tool_name": row["tool_name"],
+            "tool_input": row["tool_input"],
+            "tool_output": row["tool_output"],
+            "success": bool(row["success"]),
+            "error_message": row["error_message"],
+            "duration_ms": row["duration_ms"],
+            "file_path": row["file_path"],
+            "timestamp": ts,
+        }
+
+        # Add command analytics fields if available
+        if "command_name" in column_names:
+            activity_data["command_name"] = row["command_name"]
+        if "command_scope" in column_names:
+            activity_data["command_scope"] = row["command_scope"]
+        if "mcp_server" in column_names:
+            activity_data["mcp_server"] = row["mcp_server"]
+        if "skill_name" in column_names:
+            activity_data["skill_name"] = row["skill_name"]
+
+        # Add summary fields if available
+        if "summary" in column_names:
+            activity_data["summary"] = row["summary"]
+        if "summary_detail" in column_names:
+            activity_data["summary_detail"] = row["summary_detail"]
+
+        activities.append(Activity(**activity_data))
+
+    conn.close()
+    return activities
+
+
+def get_timeline(
+    db_path: str,
+    hours: int = 24,
+    include_memories: bool = True,
+    include_activities: bool = True,
+) -> list[TimelineEntry]:
+    """Get a timeline of memories and activities."""
+    conn = get_connection(db_path)
+    since = datetime.now() - timedelta(hours=hours)
+    since_str = since.isoformat()
+
+    entries: list[TimelineEntry] = []
+
+    if include_memories:
+        cursor = conn.execute(
+            "SELECT * FROM memories WHERE created_at >= ? ORDER BY created_at DESC",
+            (since_str,),
+        )
+        for row in cursor.fetchall():
+            entries.append(
+                TimelineEntry(
+                    timestamp=datetime.fromisoformat(row["created_at"]),
+                    entry_type="memory",
+                    data={
+                        "id": row["id"],
+                        "content": row["content"][:200] + "..." if len(row["content"]) > 200 else row["content"],
+                        "type": row["type"],
+                        "importance": row["importance_score"],
+                    },
+                )
+            )
+
+    if include_activities:
+        cursor = conn.execute(
+            "SELECT * FROM activities WHERE timestamp >= ? ORDER BY timestamp DESC",
+            (since_str,),
+        )
+        for row in cursor.fetchall():
+            entries.append(
+                TimelineEntry(
+                    timestamp=datetime.fromisoformat(row["timestamp"]),
+                    entry_type="activity",
+                    data={
+                        "id": row["id"],
+                        "event_type": row["event_type"],
+                        "tool_name": row["tool_name"],
+                        "success": bool(row["success"]),
+                        "duration_ms": row["duration_ms"],
+                    },
+                )
+            )
+
+    # Sort by timestamp descending
+    entries.sort(key=lambda e: e.timestamp, reverse=True)
+
+    conn.close()
+    return entries
+
+
+def get_sessions(db_path: str, limit: int = 20) -> list[Session]:
+    """Get recent sessions."""
+    conn = get_connection(db_path)
+
+    cursor = conn.execute(
+        """
+        SELECT s.*, COUNT(a.id) as activity_count
+        FROM sessions s
+        LEFT JOIN activities a ON s.id = a.session_id
+        GROUP BY s.id
+        ORDER BY s.started_at DESC
+        LIMIT ?
+        """,
+        (limit,),
+    )
+
+    sessions = []
+    for row in cursor.fetchall():
+        sessions.append(
+            Session(
+                id=row["id"],
+                project_path=row["project_path"],
+                started_at=datetime.fromisoformat(row["started_at"]),
+                ended_at=datetime.fromisoformat(row["ended_at"]) if row["ended_at"] else None,
+                summary=row["summary"],
+                activity_count=row["activity_count"],
+            )
+        )
+
+    conn.close()
+    return sessions
+
+
+def get_all_tags(db_path: str) -> list[dict]:
+    """Get all tags with their usage counts."""
+    conn = get_connection(db_path)
+
+    # Extract tags from JSON column
+    cursor = conn.execute("SELECT tags FROM memories WHERE tags IS NOT NULL AND tags != ''")
+    tag_counter: Counter = Counter()
+    for row in cursor.fetchall():
+        tags = parse_tags(row["tags"])
+        tag_counter.update(tags)
+
+    tags = [{"name": name, "count": count} for name, count in tag_counter.most_common()]
+
+    conn.close()
+    return tags
+
+
+def get_type_distribution(db_path: str) -> dict[str, int]:
+    """Get memory type distribution."""
+    conn = get_connection(db_path)
+
+    cursor = conn.execute("SELECT type, COUNT(*) as count FROM memories GROUP BY type")
+    distribution = {row["type"]: row["count"] for row in cursor.fetchall()}
+
+    conn.close()
+    return distribution
+
+
+def search_memories(db_path: str, query: str, limit: int = 20) -> list[Memory]:
+    """Search memories using FTS if available, otherwise LIKE."""
+    conn = get_connection(db_path)
+
+    # Check if FTS table exists
+    fts_check = conn.execute(
+        "SELECT name FROM sqlite_master WHERE type='table' AND name='memories_fts'"
+    ).fetchone()
+
+    if fts_check:
+        # Use FTS search - FTS5 uses rowid to match the memories table rowid
+        # Escape special FTS5 characters and wrap in quotes for phrase search
+        safe_query = query.replace('"', '""')
+        try:
+            cursor = conn.execute(
+                """
+                SELECT m.* FROM memories m
+                JOIN memories_fts fts ON m.rowid = fts.rowid
+                WHERE memories_fts MATCH ?
+                ORDER BY rank
+                LIMIT ?
+                """,
+                (f'"{safe_query}"', limit),
+            )
+        except sqlite3.OperationalError:
+            # Fallback if FTS query fails
+            search_term = f"%{query}%"
+            cursor = conn.execute(
+                """
+                SELECT * FROM memories
+                WHERE content LIKE ? OR context LIKE ?
+                ORDER BY importance_score DESC
+                LIMIT ?
+                """,
+                (search_term, search_term, limit),
+            )
+    else:
+        # Fallback to LIKE
+        search_term = f"%{query}%"
+        cursor = conn.execute(
+            """
+            SELECT * FROM memories
+            WHERE content LIKE ? OR context LIKE ?
+            ORDER BY importance_score DESC
+            LIMIT ?
+            """,
+            (search_term, search_term, limit),
+        )
+
+    memories = []
+    for row in cursor.fetchall():
+        # Parse tags from JSON string
+        tags = parse_tags(row["tags"])
+
+        memories.append(
+            Memory(
+                id=row["id"],
+                content=row["content"],
+                context=row["context"],
+                type=row["type"],
+                status=row["status"] or "fresh",
+                importance_score=int(row["importance_score"] or 50),
+                access_count=row["access_count"] or 0,
+                created_at=datetime.fromisoformat(row["created_at"]),
+                last_accessed=datetime.fromisoformat(row["last_accessed"]) if row["last_accessed"] else None,
+                tags=tags,
+            )
+        )
+
+    conn.close()
+    return memories
+
+
+def update_memory(db_path: str, memory_id: str, updates: MemoryUpdate) -> Optional[Memory]:
+    """Update a memory and return the updated record."""
+    conn = get_write_connection(db_path)
+
+    # Build update query dynamically based on provided fields
+    update_fields = []
+    params = []
+
+    if updates.content is not None:
+        update_fields.append("content = ?")
+        params.append(updates.content)
+
+    if updates.context is not None:
+        update_fields.append("context = ?")
+        params.append(updates.context)
+
+    if updates.memory_type is not None:
+        update_fields.append("type = ?")
+        params.append(updates.memory_type)
+
+    if updates.status is not None:
+        update_fields.append("status = ?")
+        params.append(updates.status)
+
+    if updates.importance_score is not None:
+        update_fields.append("importance_score = ?")
+        params.append(updates.importance_score)
+
+    if updates.tags is not None:
+        update_fields.append("tags = ?")
+        params.append(json.dumps(updates.tags))
+
+    if not update_fields:
+        conn.close()
+        return get_memory_by_id(db_path, memory_id)
+
+    # Add updated timestamp
+    update_fields.append("last_accessed = ?")
+    params.append(datetime.now().isoformat())
+
+    # Add memory_id to params
+    params.append(memory_id)
+
+    query = f"UPDATE memories SET {', '.join(update_fields)} WHERE id = ?"
+    cursor = conn.execute(query, params)
+    conn.commit()
+
+    if cursor.rowcount == 0:
+        conn.close()
+        return None
+
+    conn.close()
+    return get_memory_by_id(db_path, memory_id)
+
+
+def delete_memory(db_path: str, memory_id: str) -> bool:
+    """Delete a memory by ID. Returns True if deleted, False if not found."""
+    conn = get_write_connection(db_path)
+
+    # Also delete related entries in memory_relationships
+    conn.execute(
+        "DELETE FROM memory_relationships WHERE source_id = ? OR target_id = ?",
+        (memory_id, memory_id),
+    )
+
+    cursor = conn.execute("DELETE FROM memories WHERE id = ?", (memory_id,))
+    conn.commit()
+
+    deleted = cursor.rowcount > 0
+    conn.close()
+    return deleted
+
+
+# --- Stats Functions for Dashboard Charts ---
+
+
+def get_activity_heatmap(db_path: str, days: int = 90) -> list[dict]:
+    """Get activity counts grouped by day for heatmap visualization."""
+    conn = get_connection(db_path)
+    query = """
+        SELECT date(timestamp) as date, COUNT(*) as count
+        FROM activities
+        WHERE timestamp >= date('now', ?)
+        GROUP BY date(timestamp)
+        ORDER BY date
+    """
+    cursor = conn.execute(query, (f'-{days} days',))
+    result = [{"date": row["date"], "count": row["count"]} for row in cursor.fetchall()]
+    conn.close()
+    return result
+
+
+def get_tool_usage(db_path: str, limit: int = 10) -> list[dict]:
+    """Get tool usage statistics with success rates."""
+    conn = get_connection(db_path)
+    query = """
+        SELECT
+            tool_name,
+            COUNT(*) as count,
+            SUM(CASE WHEN success = 1 THEN 1 ELSE 0 END) * 1.0 / COUNT(*) as success_rate
+        FROM activities
+        WHERE tool_name IS NOT NULL AND tool_name != ''
+        GROUP BY tool_name
+        ORDER BY count DESC
+        LIMIT ?
+    """
+    cursor = conn.execute(query, (limit,))
+    result = [
+        {
+            "tool_name": row["tool_name"],
+            "count": row["count"],
+            "success_rate": round(row["success_rate"], 2) if row["success_rate"] else 1.0,
+        }
+        for row in cursor.fetchall()
+    ]
+    conn.close()
+    return result
+
+
+def get_memory_growth(db_path: str, days: int = 30) -> list[dict]:
+    """Get memory creation over time with cumulative totals."""
+    conn = get_connection(db_path)
+    query = """
+        WITH daily_counts AS (
+            SELECT date(created_at) as date, COUNT(*) as count
+            FROM memories
+            WHERE created_at >= date('now', ?)
+            GROUP BY date(created_at)
+        )
+        SELECT
+            date,
+            count,
+            SUM(count) OVER (ORDER BY date) as cumulative
+        FROM daily_counts
+        ORDER BY date
+    """
+    cursor = conn.execute(query, (f'-{days} days',))
+    result = [
+        {"date": row["date"], "count": row["count"], "cumulative": row["cumulative"]}
+        for row in cursor.fetchall()
+    ]
+    conn.close()
+    return result
+
+
+def get_recent_sessions(db_path: str, limit: int = 5) -> list[dict]:
+    """Get recent sessions with activity counts and memory counts."""
+    conn = get_connection(db_path)
+    query = """
+        SELECT
+            s.id,
+            s.project_path,
+            s.started_at,
+            s.ended_at,
+            s.summary,
+            COUNT(DISTINCT a.id) as activity_count
+        FROM sessions s
+        LEFT JOIN activities a ON a.session_id = s.id
+        GROUP BY s.id
+        ORDER BY s.started_at DESC
+        LIMIT ?
+    """
+    cursor = conn.execute(query, (limit,))
+    result = [
+        {
+            "id": row["id"],
+            "project_path": row["project_path"],
+            "started_at": row["started_at"],
+            "ended_at": row["ended_at"],
+            "summary": row["summary"],
+            "activity_count": row["activity_count"],
+        }
+        for row in cursor.fetchall()
+    ]
+    conn.close()
+    return result
+
+
+def bulk_update_memory_status(db_path: str, memory_ids: list[str], status: str) -> int:
+    """Update status for multiple memories. Returns count updated."""
+    if not memory_ids:
+        return 0
+    conn = get_write_connection(db_path)
+    placeholders = ','.join('?' * len(memory_ids))
+    query = f"UPDATE memories SET status = ?, last_accessed = datetime('now') WHERE id IN ({placeholders})"
+    cursor = conn.execute(query, [status] + memory_ids)
+    conn.commit()
+    count = cursor.rowcount
+    conn.close()
+    return count
+
+
+def get_memories_needing_review(db_path: str, days_threshold: int = 30, limit: int = 50) -> list[Memory]:
+    """Get memories that haven't been accessed recently and may need review."""
+    conn = get_connection(db_path)
+    query = """
+        SELECT * FROM memories
+        WHERE last_accessed < date('now', ?)
+           OR last_accessed IS NULL
+        ORDER BY last_accessed ASC NULLS FIRST, importance_score DESC
+        LIMIT ?
+    """
+    cursor = conn.execute(query, (f'-{days_threshold} days', limit))
+
+    memories = []
+    for row in cursor.fetchall():
+        tags = parse_tags(row["tags"])
+        memories.append(
+            Memory(
+                id=row["id"],
+                content=row["content"],
+                context=row["context"],
+                type=row["type"],
+                status=row["status"] or "fresh",
+                importance_score=int(row["importance_score"] or 50),
+                access_count=row["access_count"] or 0,
+                created_at=datetime.fromisoformat(row["created_at"]),
+                last_accessed=datetime.fromisoformat(row["last_accessed"]) if row["last_accessed"] else None,
+                tags=tags,
+            )
+        )
+
+    conn.close()
+    return memories
+
+
+def get_relationships(db_path: str, memory_id: Optional[str] = None) -> list[dict]:
+    """Get memory relationships for graph visualization."""
+    conn = get_connection(db_path)
+
+    # Check if memory_relationships table exists
+    table_check = conn.execute(
+        "SELECT name FROM sqlite_master WHERE type='table' AND name='memory_relationships'"
+    ).fetchone()
+
+    if not table_check:
+        conn.close()
+        return []
+
+    query = """
+        SELECT
+            r.source_memory_id as source_id,
+            r.target_memory_id as target_id,
+            r.relationship_type,
+            r.strength,
+            ms.content as source_content,
+            ms.type as source_type,
+            mt.content as target_content,
+            mt.type as target_type
+        FROM memory_relationships r
+        JOIN memories ms ON r.source_memory_id = ms.id
+        JOIN memories mt ON r.target_memory_id = mt.id
+    """
+
+    try:
+        if memory_id:
+            query += " WHERE r.source_memory_id = ? OR r.target_memory_id = ?"
+            cursor = conn.execute(query, (memory_id, memory_id))
+        else:
+            cursor = conn.execute(query)
+
+        result = [dict(row) for row in cursor.fetchall()]
+    except Exception as e:
+        print(f"[Database] Error querying relationships: {e}")
+        result = []
+    finally:
+        conn.close()
+
+    return result
+
+
+def get_relationship_graph(db_path: str, center_id: Optional[str] = None, depth: int = 2) -> dict:
+    """Get graph data with nodes and edges for D3 visualization."""
+    relationships = get_relationships(db_path, center_id)
+
+    nodes = {}
+    edges = []
+
+    for rel in relationships:
+        # Add source node
+        if rel["source_id"] not in nodes:
+            nodes[rel["source_id"]] = {
+                "id": rel["source_id"],
+                "content": rel["source_content"][:100] if rel["source_content"] else "",
+                "type": rel["source_type"],
+            }
+        # Add target node
+        if rel["target_id"] not in nodes:
+            nodes[rel["target_id"]] = {
+                "id": rel["target_id"],
+                "content": rel["target_content"][:100] if rel["target_content"] else "",
+                "type": rel["target_type"],
+            }
+        # Add edge
+        edges.append({
+            "source": rel["source_id"],
+            "target": rel["target_id"],
+            "type": rel["relationship_type"],
+            "strength": rel["strength"] or 1.0,
+        })
+
+    return {"nodes": list(nodes.values()), "edges": edges}
+
+
+# --- Command Analytics Functions ---
+
+
+def get_command_usage(db_path: str, scope: Optional[str] = None, days: int = 30) -> list[dict]:
+    """Get slash command usage statistics aggregated by command_name.
+
+    Args:
+        db_path: Path to database
+        scope: Filter by scope ('universal', 'project', or None for all)
+        days: Number of days to look back
+
+    Returns:
+        List of command usage entries with counts and success rates
+    """
+    conn = get_connection(db_path)
+
+    # Check if command_name column exists
+    columns = conn.execute("PRAGMA table_info(activities)").fetchall()
+    column_names = [col[1] for col in columns]
+    if "command_name" not in column_names:
+        conn.close()
+        return []
+
+    query = """
+        SELECT
+            command_name,
+            command_scope,
+            COUNT(*) as count,
+            SUM(CASE WHEN success = 1 THEN 1 ELSE 0 END) * 1.0 / COUNT(*) as success_rate,
+            AVG(duration_ms) as avg_duration_ms
+        FROM activities
+        WHERE command_name IS NOT NULL
+          AND command_name != ''
+          AND timestamp >= date('now', ?)
+    """
+    params = [f'-{days} days']
+
+    if scope:
+        query += " AND command_scope = ?"
+        params.append(scope)
+
+    query += " GROUP BY command_name, command_scope ORDER BY count DESC"
+
+    cursor = conn.execute(query, params)
+    result = [
+        {
+            "command_name": row["command_name"],
+            "command_scope": row["command_scope"] or "unknown",
+            "count": row["count"],
+            "success_rate": round(row["success_rate"], 2) if row["success_rate"] else 1.0,
+            "avg_duration_ms": round(row["avg_duration_ms"]) if row["avg_duration_ms"] else None,
+        }
+        for row in cursor.fetchall()
+    ]
+    conn.close()
+    return result
+
+
+def get_skill_usage(db_path: str, scope: Optional[str] = None, days: int = 30) -> list[dict]:
+    """Get skill usage statistics aggregated by skill_name.
+
+    Args:
+        db_path: Path to database
+        scope: Filter by scope ('universal', 'project', or None for all)
+        days: Number of days to look back
+
+    Returns:
+        List of skill usage entries with counts and success rates
+    """
+    conn = get_connection(db_path)
+
+    # Check if skill_name column exists
+    columns = conn.execute("PRAGMA table_info(activities)").fetchall()
+    column_names = [col[1] for col in columns]
+    if "skill_name" not in column_names:
+        conn.close()
+        return []
+
+    query = """
+        SELECT
+            skill_name,
+            command_scope,
+            COUNT(*) as count,
+            SUM(CASE WHEN success = 1 THEN 1 ELSE 0 END) * 1.0 / COUNT(*) as success_rate,
+            AVG(duration_ms) as avg_duration_ms
+        FROM activities
+        WHERE skill_name IS NOT NULL
+          AND skill_name != ''
+          AND timestamp >= date('now', ?)
+    """
+    params = [f'-{days} days']
+
+    if scope:
+        query += " AND command_scope = ?"
+        params.append(scope)
+
+    query += " GROUP BY skill_name, command_scope ORDER BY count DESC"
+
+    cursor = conn.execute(query, params)
+    result = [
+        {
+            "skill_name": row["skill_name"],
+            "skill_scope": row["command_scope"] or "unknown",
+            "count": row["count"],
+            "success_rate": round(row["success_rate"], 2) if row["success_rate"] else 1.0,
+            "avg_duration_ms": round(row["avg_duration_ms"]) if row["avg_duration_ms"] else None,
+        }
+        for row in cursor.fetchall()
+    ]
+    conn.close()
+    return result
+
+
+def get_mcp_usage(db_path: str, days: int = 30) -> list[dict]:
+    """Get MCP server usage statistics.
+
+    Args:
+        db_path: Path to database
+        days: Number of days to look back
+
+    Returns:
+        List of MCP server usage entries with tool counts and call totals
+    """
+    conn = get_connection(db_path)
+
+    # Check if mcp_server column exists
+    columns = conn.execute("PRAGMA table_info(activities)").fetchall()
+    column_names = [col[1] for col in columns]
+    if "mcp_server" not in column_names:
+        conn.close()
+        return []
+
+    query = """
+        SELECT
+            mcp_server,
+            COUNT(DISTINCT tool_name) as tool_count,
+            COUNT(*) as total_calls,
+            SUM(CASE WHEN success = 1 THEN 1 ELSE 0 END) * 1.0 / COUNT(*) as success_rate
+        FROM activities
+        WHERE mcp_server IS NOT NULL
+          AND mcp_server != ''
+          AND timestamp >= date('now', ?)
+        GROUP BY mcp_server
+        ORDER BY total_calls DESC
+    """
+    cursor = conn.execute(query, (f'-{days} days',))
+    result = [
+        {
+            "mcp_server": row["mcp_server"],
+            "tool_count": row["tool_count"],
+            "total_calls": row["total_calls"],
+            "success_rate": round(row["success_rate"], 2) if row["success_rate"] else 1.0,
+        }
+        for row in cursor.fetchall()
+    ]
+    conn.close()
+    return result
+
+
+def get_activity_detail(db_path: str, activity_id: str) -> Optional[dict]:
+    """Get full activity details including complete input/output.
+
+    Args:
+        db_path: Path to database
+        activity_id: Activity ID
+
+    Returns:
+        Full activity details or None if not found
+    """
+    conn = get_connection(db_path)
+    cursor = conn.execute("SELECT * FROM activities WHERE id = ?", (activity_id,))
+    row = cursor.fetchone()
+
+    if not row:
+        conn.close()
+        return None
+
+    # Get column names for safe access
+    column_names = [description[0] for description in cursor.description]
+
+    result = {
+        "id": row["id"],
+        "session_id": row["session_id"],
+        "event_type": row["event_type"],
+        "tool_name": row["tool_name"],
+        "tool_input_full": row["tool_input"],
+        "tool_output_full": row["tool_output"],
+        "success": bool(row["success"]),
+        "error_message": row["error_message"],
+        "duration_ms": row["duration_ms"],
+        "file_path": row["file_path"],
+        "timestamp": row["timestamp"],
+    }
+
+    # Add command analytics fields if they exist
+    if "command_name" in column_names:
+        result["command_name"] = row["command_name"]
+    if "command_scope" in column_names:
+        result["command_scope"] = row["command_scope"]
+    if "mcp_server" in column_names:
+        result["mcp_server"] = row["mcp_server"]
+    if "skill_name" in column_names:
+        result["skill_name"] = row["skill_name"]
+
+    # Add summary fields if they exist
+    if "summary" in column_names:
+        result["summary"] = row["summary"]
+    if "summary_detail" in column_names:
+        result["summary_detail"] = row["summary_detail"]
+
+    conn.close()
+    return result
+
+
+def create_memory(
+    db_path: str,
+    content: str,
+    memory_type: str = "other",
+    context: Optional[str] = None,
+    tags: Optional[list[str]] = None,
+    importance_score: int = 50,
+    related_memory_ids: Optional[list[str]] = None,
+) -> str:
+    """Create a new memory and return its ID.
+
+    Args:
+        db_path: Path to the database file
+        content: Memory content
+        memory_type: Type of memory (e.g., 'decision', 'solution', 'conversation')
+        context: Additional context
+        tags: List of tags
+        importance_score: Importance score (1-100)
+        related_memory_ids: IDs of related memories to create relationships with
+
+    Returns:
+        The ID of the created memory
+    """
+    import uuid
+
+    conn = get_write_connection(db_path)
+
+    # Generate ID
+    memory_id = f"mem_{int(datetime.now().timestamp() * 1000)}_{uuid.uuid4().hex[:8]}"
+    now = datetime.now().isoformat()
+
+    # Insert memory
+    conn.execute(
+        """
+        INSERT INTO memories (id, content, context, type, status, importance_score, access_count, created_at, last_accessed, updated_at, tags)
+        VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
+        """,
+        (
+            memory_id,
+            content,
+            context,
+            memory_type,
+            "fresh",
+            importance_score,
+            0,
+            now,
+            now,
+            now,
+            json.dumps(tags) if tags else None,
+        ),
+    )
+
+    # Create relationships if related_memory_ids provided
+    if related_memory_ids:
+        # Check if memory_relationships table exists
+        table_check = conn.execute(
+            "SELECT name FROM sqlite_master WHERE type='table' AND name='memory_relationships'"
+        ).fetchone()
+
+        if table_check:
+            for related_id in related_memory_ids:
+                try:
+                    conn.execute(
+                        """
+                        INSERT INTO memory_relationships (source_memory_id, target_memory_id, relationship_type, strength)
+                        VALUES (?, ?, ?, ?)
+                        """,
+                        (memory_id, related_id, "derived_from", 0.8),
+                    )
+                except Exception:
+                    # Ignore if related memory doesn't exist
+                    pass
+
+    conn.commit()
+    conn.close()
+
+    return memory_id
+
+
+# --- User Message Functions for Style Tab ---
+
+
+def get_user_messages(
+    db_path: str,
+    session_id: Optional[str] = None,
+    search: Optional[str] = None,
+    has_code_blocks: Optional[bool] = None,
+    has_questions: Optional[bool] = None,
+    has_commands: Optional[bool] = None,
+    tone_filter: Optional[str] = None,
+    sort_by: str = "timestamp",
+    sort_order: str = "desc",
+    limit: int = 50,
+    offset: int = 0,
+) -> list[dict]:
+    """Get user messages with filtering, sorting, and pagination.
+
+    Args:
+        db_path: Path to database
+        session_id: Filter by session
+        search: Search in content
+        has_code_blocks: Filter messages with/without code blocks
+        has_questions: Filter messages with/without questions
+        has_commands: Filter messages with/without slash commands
+        tone_filter: Filter by tone indicator (e.g., 'polite', 'urgent', 'technical')
+        sort_by: Sort by column (timestamp, word_count, char_count)
+        sort_order: 'asc' or 'desc'
+        limit: Maximum results
+        offset: Pagination offset
+
+    Returns:
+        List of user message dictionaries
+    """
+    conn = get_connection(db_path)
+
+    # Check if user_messages table exists
+    table_check = conn.execute(
+        "SELECT name FROM sqlite_master WHERE type='table' AND name='user_messages'"
+    ).fetchone()
+
+    if not table_check:
+        conn.close()
+        return []
+
+    query = "SELECT * FROM user_messages WHERE 1=1"
+    params: list = []
+
+    if session_id:
+        query += " AND session_id = ?"
+        params.append(session_id)
+
+    if search:
+        query += " AND content LIKE ?"
+        params.append(f"%{search}%")
+
+    if has_code_blocks is not None:
+        query += " AND has_code_blocks = ?"
+        params.append(1 if has_code_blocks else 0)
+
+    if has_questions is not None:
+        query += " AND has_questions = ?"
+        params.append(1 if has_questions else 0)
+
+    if has_commands is not None:
+        query += " AND has_commands = ?"
+        params.append(1 if has_commands else 0)
+
+    if tone_filter:
+        # Search within JSON array of tone_indicators
+        query += " AND tone_indicators LIKE ?"
+        params.append(f'%"{tone_filter}"%')
+
+    # Sorting
+    valid_sort_columns = ["timestamp", "word_count", "char_count", "line_count"]
+    sort_by = sort_by if sort_by in valid_sort_columns else "timestamp"
+    sort_order = "DESC" if sort_order.lower() == "desc" else "ASC"
+    query += f" ORDER BY {sort_by} {sort_order}"
+
+    # Pagination
+    query += " LIMIT ? OFFSET ?"
+    params.extend([limit, offset])
+
+    cursor = conn.execute(query, params)
+    messages = []
+
+    for row in cursor.fetchall():
+        # Parse tone_indicators from JSON
+        tone_indicators = []
+        if row["tone_indicators"]:
+            try:
+                tone_indicators = json.loads(row["tone_indicators"])
+            except (json.JSONDecodeError, TypeError):
+                pass
+
+        # Get primary tone (first in the list) for frontend compatibility
+        primary_tone = tone_indicators[0] if tone_indicators else None
+
+        messages.append({
+            "id": row["id"],
+            "session_id": row["session_id"],
+            "created_at": row["timestamp"],  # Frontend expects created_at
+            "timestamp": row["timestamp"],   # Keep for backward compatibility
+            "content": row["content"],
+            "word_count": row["word_count"],
+            "char_count": row["char_count"],
+            "line_count": row["line_count"],
+            "has_code_blocks": bool(row["has_code_blocks"]),
+            "has_questions": bool(row["has_questions"]),
+            "has_commands": bool(row["has_commands"]),
+            "tone": primary_tone,            # Frontend expects single tone string
+            "tone_indicators": tone_indicators,
+            "project_path": row["project_path"],
+        })
+
+    conn.close()
+    return messages
+
+
+def get_user_message_count(
+    db_path: str,
+    session_id: Optional[str] = None,
+) -> int:
+    """Get total count of user messages.
+
+    Args:
+        db_path: Path to database
+        session_id: Optional filter by session
+
+    Returns:
+        Count of user messages
+    """
+    conn = get_connection(db_path)
+
+    # Check if user_messages table exists
+    table_check = conn.execute(
+        "SELECT name FROM sqlite_master WHERE type='table' AND name='user_messages'"
+    ).fetchone()
+
+    if not table_check:
+        conn.close()
+        return 0
+
+    query = "SELECT COUNT(*) FROM user_messages"
+    params = []
+
+    if session_id:
+        query += " WHERE session_id = ?"
+        params.append(session_id)
+
+    count = conn.execute(query, params).fetchone()[0]
+    conn.close()
+    return count
+
+
+def delete_user_message(db_path: str, message_id: str) -> bool:
+    """Delete a single user message.
+
+    Args:
+        db_path: Path to database
+        message_id: Message ID to delete
+
+    Returns:
+        True if deleted, False if not found
+    """
+    conn = get_write_connection(db_path)
+
+    cursor = conn.execute("DELETE FROM user_messages WHERE id = ?", (message_id,))
+    conn.commit()
+
+    deleted = cursor.rowcount > 0
+    conn.close()
+    return deleted
+
+
+def delete_user_messages_bulk(db_path: str, message_ids: list[str]) -> int:
+    """Delete multiple user messages.
+
+    Args:
+        db_path: Path to database
+        message_ids: List of message IDs to delete
+
+    Returns:
+        Count of messages deleted
+    """
+    if not message_ids:
+        return 0
+
+    conn = get_write_connection(db_path)
+    placeholders = ','.join('?' * len(message_ids))
+    query = f"DELETE FROM user_messages WHERE id IN ({placeholders})"
+    cursor = conn.execute(query, message_ids)
+    conn.commit()
+
+    count = cursor.rowcount
+    conn.close()
+    return count
+
+
+def get_style_profile(db_path: str, project_path: Optional[str] = None) -> Optional[dict]:
+    """Get user style profile.
+
+    Args:
+        db_path: Path to database
+        project_path: Project-specific profile, or None for global
+
+    Returns:
+        Style profile dictionary or None if not found
+    """
+    conn = get_connection(db_path)
+
+    # Check if user_style_profiles table exists
+    table_check = conn.execute(
+        "SELECT name FROM sqlite_master WHERE type='table' AND name='user_style_profiles'"
+    ).fetchone()
+
+    if not table_check:
+        conn.close()
+        return None
+
+    if project_path:
+        query = "SELECT * FROM user_style_profiles WHERE project_path = ? ORDER BY updated_at DESC LIMIT 1"
+        cursor = conn.execute(query, (project_path,))
+    else:
+        query = "SELECT * FROM user_style_profiles WHERE project_path IS NULL ORDER BY updated_at DESC LIMIT 1"
+        cursor = conn.execute(query)
+
+    row = cursor.fetchone()
+    conn.close()
+
+    if not row:
+        return None
+
+    # Parse JSON fields
+    def parse_json_field(value):
+        if not value:
+            return None
+        try:
+            return json.loads(value)
+        except (json.JSONDecodeError, TypeError):
+            return None
+
+    return {
+        "id": row["id"],
+        "project_path": row["project_path"],
+        "total_messages": row["total_messages"],
+        "avg_word_count": row["avg_word_count"],
+        "avg_char_count": row["avg_char_count"],
+        "common_phrases": parse_json_field(row["common_phrases"]),
+        "vocabulary_richness": row["vocabulary_richness"],
+        "formality_score": row["formality_score"],
+        "question_frequency": row["question_frequency"],
+        "command_frequency": row["command_frequency"],
+        "code_block_frequency": row["code_block_frequency"],
+        "punctuation_style": parse_json_field(row["punctuation_style"]),
+        "greeting_patterns": parse_json_field(row["greeting_patterns"]),
+        "instruction_style": parse_json_field(row["instruction_style"]),
+        "sample_messages": parse_json_field(row["sample_messages"]),
+        "created_at": row["created_at"],
+        "updated_at": row["updated_at"],
+    }
+
+
+def get_style_samples(db_path: str, limit: int = 10) -> list[dict]:
+    """Get sample user messages for style analysis preview.
+
+    Returns a diverse selection of messages showcasing different styles.
+
+    Args:
+        db_path: Path to database
+        limit: Maximum samples to return
+
+    Returns:
+        List of sample messages with style indicators
+    """
+    conn = get_connection(db_path)
+
+    # Check if user_messages table exists
+    table_check = conn.execute(
+        "SELECT name FROM sqlite_master WHERE type='table' AND name='user_messages'"
+    ).fetchone()
+
+    if not table_check:
+        conn.close()
+        return []
+
+    # Get a diverse sample: some recent, some with code, some with questions
+    samples = []
+
+    # Recent messages
+    cursor = conn.execute(
+        "SELECT * FROM user_messages ORDER BY timestamp DESC LIMIT ?",
+        (limit // 3,)
+    )
+    for row in cursor.fetchall():
+        samples.append(_row_to_sample(row))
+
+    # Messages with code blocks
+    cursor = conn.execute(
+        "SELECT * FROM user_messages WHERE has_code_blocks = 1 ORDER BY timestamp DESC LIMIT ?",
+        (limit // 3,)
+    )
+    for row in cursor.fetchall():
+        sample = _row_to_sample(row)
+        if sample["id"] not in [s["id"] for s in samples]:
+            samples.append(sample)
+
+    # Longer messages (likely more substantive)
+    cursor = conn.execute(
+        "SELECT * FROM user_messages WHERE word_count > 20 ORDER BY word_count DESC LIMIT ?",
+        (limit // 3,)
+    )
+    for row in cursor.fetchall():
+        sample = _row_to_sample(row)
+        if sample["id"] not in [s["id"] for s in samples]:
+            samples.append(sample)
+
+    conn.close()
+    return samples[:limit]
+
+
+def _row_to_sample(row) -> dict:
+    """Convert a database row to a sample message dict."""
+    tone_indicators = []
+    if row["tone_indicators"]:
+        try:
+            tone_indicators = json.loads(row["tone_indicators"])
+        except (json.JSONDecodeError, TypeError):
+            pass
+
+    return {
+        "id": row["id"],
+        "timestamp": row["timestamp"],
+        "content_preview": row["content"][:200] + "..." if len(row["content"]) > 200 else row["content"],
+        "word_count": row["word_count"],
+        "has_code_blocks": bool(row["has_code_blocks"]),
+        "has_questions": bool(row["has_questions"]),
+        "tone_indicators": tone_indicators,
+    }
+
+
+def get_style_samples_by_category(db_path: str, samples_per_tone: int = 3) -> dict:
+    """Get sample user messages grouped by style category.
+
+    Maps tone_indicators to frontend categories:
+    - professional: direct, polite, formal tones
+    - casual: casual tones
+    - technical: technical tones
+    - creative: unique patterns, inquisitive tones
+
+    Args:
+        db_path: Path to database
+        samples_per_tone: Max samples per category
+
+    Returns:
+        Dict with professional, casual, technical, creative lists
+    """
+    conn = get_connection(db_path)
+
+    # Check if user_messages table exists
+    table_check = conn.execute(
+        "SELECT name FROM sqlite_master WHERE type='table' AND name='user_messages'"
+    ).fetchone()
+
+    if not table_check:
+        conn.close()
+        return {
+            "professional": [],
+            "casual": [],
+            "technical": [],
+            "creative": []
+        }
+
+    result = {
+        "professional": [],
+        "casual": [],
+        "technical": [],
+        "creative": []
+    }
+
+    # Mapping from tone_indicators to categories
+    tone_to_category = {
+        "direct": "professional",
+        "polite": "professional",
+        "formal": "professional",
+        "casual": "casual",
+        "technical": "technical",
+        "inquisitive": "creative",
+        "urgent": "professional",
+    }
+
+    # Get all messages with tone indicators
+    cursor = conn.execute(
+        """SELECT content, tone_indicators FROM user_messages
+           WHERE tone_indicators IS NOT NULL AND tone_indicators != '[]'
+           ORDER BY timestamp DESC LIMIT 200"""
+    )
+
+    for row in cursor.fetchall():
+        content = row["content"]
+        try:
+            tones = json.loads(row["tone_indicators"]) if row["tone_indicators"] else []
+        except (json.JSONDecodeError, TypeError):
+            tones = []
+
+        # Map to categories
+        for tone in tones:
+            category = tone_to_category.get(tone.lower(), "creative")
+            if len(result[category]) < samples_per_tone:
+                # Truncate content for preview
+                preview = content[:200] + "..." if len(content) > 200 else content
+                if preview not in result[category]:
+                    result[category].append(preview)
+                    break  # Only add to first matching category
+
+    # Fill any empty categories with recent messages
+    if any(len(v) == 0 for v in result.values()):
+        cursor = conn.execute(
+            "SELECT content FROM user_messages ORDER BY timestamp DESC LIMIT ?",
+            (samples_per_tone * 4,)
+        )
+        fallback_messages = [
+            row["content"][:200] + "..." if len(row["content"]) > 200 else row["content"]
+            for row in cursor.fetchall()
+        ]
+
+        for category in result:
+            if len(result[category]) == 0 and fallback_messages:
+                # Take messages for empty categories
+                for msg in fallback_messages[:samples_per_tone]:
+                    if msg not in [m for v in result.values() for m in v]:
+                        result[category].append(msg)
+
+    conn.close()
+    return result
+
+
+def compute_style_profile_from_messages(db_path: str) -> Optional[dict]:
+    """Compute a style profile from user_messages table.
+
+    This is used when no pre-computed profile exists.
+
+    Returns format expected by frontend StyleProfileCard:
+    - total_messages: int
+    - avg_word_count: float
+    - primary_tone: str
+    - question_percentage: float
+    - tone_distribution: dict[str, int]
+    - style_markers: list[str]
+    """
+    conn = get_connection(db_path)
+
+    # Check if user_messages table exists
+    table_check = conn.execute(
+        "SELECT name FROM sqlite_master WHERE type='table' AND name='user_messages'"
+    ).fetchone()
+
+    if not table_check:
+        conn.close()
+        return None
+
+    # Get total count and averages
+    stats = conn.execute(
+        """SELECT
+           COUNT(*) as total,
+           AVG(word_count) as avg_words,
+           AVG(char_count) as avg_chars,
+           SUM(CASE WHEN has_questions = 1 THEN 1 ELSE 0 END) as question_count
+           FROM user_messages"""
+    ).fetchone()
+
+    if not stats or stats["total"] == 0:
+        conn.close()
+        return None
+
+    total_messages = stats["total"]
+    avg_word_count = stats["avg_words"] or 0
+    question_percentage = (stats["question_count"] / total_messages * 100) if total_messages > 0 else 0
+
+    # Compute tone distribution
+    tone_distribution = {}
+    cursor = conn.execute(
+        "SELECT tone_indicators FROM user_messages WHERE tone_indicators IS NOT NULL AND tone_indicators != '[]'"
+    )
+    for row in cursor.fetchall():
+        try:
+            tones = json.loads(row["tone_indicators"]) if row["tone_indicators"] else []
+            for tone in tones:
+                tone_lower = tone.lower()
+                tone_distribution[tone_lower] = tone_distribution.get(tone_lower, 0) + 1
+        except (json.JSONDecodeError, TypeError):
+            pass
+
+    # Determine primary tone (most common)
+    primary_tone = "direct"
+    if tone_distribution:
+        primary_tone = max(tone_distribution, key=tone_distribution.get)
+
+    # Generate style markers based on the data
+    style_markers = []
+
+    if avg_word_count < 15:
+        style_markers.append("Concise")
+    elif avg_word_count > 40:
+        style_markers.append("Detailed")
+    else:
+        style_markers.append("Balanced length")
+
+    if question_percentage > 40:
+        style_markers.append("Question-driven")
+    elif question_percentage < 10:
+        style_markers.append("Statement-focused")
+
+    # Check for code usage
+    code_stats = conn.execute(
+        "SELECT SUM(CASE WHEN has_code_blocks = 1 THEN 1 ELSE 0 END) * 100.0 / COUNT(*) as code_pct FROM user_messages"
+    ).fetchone()
+    if code_stats and code_stats["code_pct"] and code_stats["code_pct"] > 20:
+        style_markers.append("Code-heavy")
+
+    # Add primary tone to markers
+    tone_labels = {
+        "direct": "Direct",
+        "polite": "Polite",
+        "technical": "Technical",
+        "casual": "Casual",
+        "inquisitive": "Inquisitive",
+        "urgent": "Urgent",
+    }
+    if primary_tone in tone_labels:
+        style_markers.append(tone_labels[primary_tone])
+
+    if not style_markers:
+        style_markers.append("Building profile...")
+
+    # Get sample messages to show the AI how the user actually writes
+    sample_messages = []
+    cursor = conn.execute(
+        """SELECT content FROM user_messages
+           WHERE length(content) > 20 AND length(content) < 500
+           AND has_commands = 0
+           ORDER BY timestamp DESC LIMIT 5"""
+    )
+    for row in cursor.fetchall():
+        sample_messages.append(row["content"])
+
+    conn.close()
+
+    return {
+        "totalMessages": total_messages,
+        "avgWordCount": round(avg_word_count, 1),
+        "primaryTone": primary_tone,
+        "questionPercentage": round(question_percentage, 1),
+        "toneDistribution": tone_distribution,
+        "styleMarkers": style_markers,
+        "sampleMessages": sample_messages,
+    }
+
+
+# --- Agent Query Functions ---
+
+
+def get_agents(
+    db_path: str,
+    agent_type: Optional[str] = None,
+    limit: int = 50,
+    active_only: bool = False
+) -> list[dict]:
+    """Get agents with recent activity counts."""
+    conn = get_connection(db_path)
+
+    query = """
+        SELECT
+            a.*,
+            COALESCE(recent.count, 0) as recent_activity_count,
+            CASE WHEN a.last_seen > datetime('now', '-5 minutes') THEN 1 ELSE 0 END as is_active
+        FROM agents a
+        LEFT JOIN (
+            SELECT agent_id, COUNT(*) as count
+            FROM activities
+            WHERE timestamp > datetime('now', '-1 hour')
+            GROUP BY agent_id
+        ) recent ON recent.agent_id = a.id
+        WHERE 1=1
+    """
+    params = []
+
+    if agent_type:
+        query += " AND a.type = ?"
+        params.append(agent_type)
+
+    if active_only:
+        query += " AND a.last_seen > datetime('now', '-5 minutes')"
+
+    query += " ORDER BY a.last_seen DESC LIMIT ?"
+    params.append(limit)
+
+    cursor = conn.execute(query, params)
+    results = [dict(row) for row in cursor.fetchall()]
+    conn.close()
+    return results
+
+
+def get_agent_by_id(db_path: str, agent_id: str) -> Optional[dict]:
+    """Get single agent by ID."""
+    conn = get_connection(db_path)
+
+    cursor = conn.execute("""
+        SELECT
+            a.*,
+            CASE WHEN a.last_seen > datetime('now', '-5 minutes') THEN 1 ELSE 0 END as is_active
+        FROM agents a
+        WHERE a.id = ?
+    """, (agent_id,))
+
+    row = cursor.fetchone()
+    conn.close()
+    return dict(row) if row else None
+
+
+def get_agent_tool_breakdown(db_path: str, agent_id: str) -> list[dict]:
+    """Get tool usage breakdown for an agent."""
+    conn = get_connection(db_path)
+
+    cursor = conn.execute("""
+        SELECT
+            tool_name,
+            COUNT(*) as count,
+            AVG(duration_ms) as avg_duration_ms,
+            SUM(CASE WHEN success = 1 THEN 1 ELSE 0 END) * 100.0 / COUNT(*) as success_rate
+        FROM activities
+        WHERE agent_id = ? AND tool_name IS NOT NULL
+        GROUP BY tool_name
+        ORDER BY count DESC
+    """, (agent_id,))
+
+    results = [dict(row) for row in cursor.fetchall()]
+    conn.close()
+    return results
+
+
+def get_agent_files_touched(db_path: str, agent_id: str, limit: int = 50) -> list[str]:
+    """Get list of files an agent has touched."""
+    conn = get_connection(db_path)
+
+    # Files from file_path column
+    cursor = conn.execute("""
+        SELECT DISTINCT file_path
+        FROM activities
+        WHERE agent_id = ?
+          AND file_path IS NOT NULL
+          AND file_path != ''
+        LIMIT ?
+    """, (agent_id, limit))
+
+    files = [row[0] for row in cursor.fetchall()]
+    conn.close()
+    return files
+
+
+def get_agent_parent(db_path: str, agent_id: str) -> Optional[str]:
+    """Find which agent spawned this subagent (via Task tool)."""
+    conn = get_connection(db_path)
+
+    # Look for Task tool call that created this agent
+    cursor = conn.execute("""
+        SELECT agent_id
+        FROM activities
+        WHERE tool_name = 'Task'
+          AND tool_output LIKE ?
+        ORDER BY timestamp DESC
+        LIMIT 1
+    """, (f'%{agent_id}%',))
+
+    row = cursor.fetchone()
+    conn.close()
+    return row[0] if row else None