[B2b] Add search analytics table + log every brain_search call

[four-bytes-robby]

Purpose

Track search effectiveness: log query → result count → top score for every brain_search call. Foundation for B2a/B2c stats.

Design

Add search_analytics table to src/schema.ts:

CREATE TABLE IF NOT EXISTS search_analytics (
  id TEXT PRIMARY KEY,
  session_id TEXT NOT NULL,
  query TEXT NOT NULL,
  result_count INTEGER NOT NULL DEFAULT 0,
  top_score REAL DEFAULT 0,
  content_types TEXT DEFAULT '',
  duration_ms INTEGER DEFAULT 0,
  helpful INTEGER,  -- NULL=unreported, 0=no, 1=yes
  used_results INTEGER,  -- NULL=unreported, 0=no, 1=yes
  created_at TEXT NOT NULL DEFAULT (datetime('now'))
);
CREATE INDEX IF NOT EXISTS idx_search_analytics_session ON search_analytics(session_id, created_at);

Hook into brain_search tool execution in src/four-opencode-brain.ts:

• After search completes, INSERT a row with query, result_count, top_score, content_types, duration
• Generate UUID7 for id
• Use session-scoped session_id

Add aggregation function getSearchStats(sessionId?) returning:

• total_searches, avg_results, avg_top_score, most_common_content_type

Acceptance

• search_analytics table created on schema init
• Every brain_search call logged with full metrics
• getSearchStats() returns correct aggregates
• Tests: verify table creation, search logging, aggregation