fix: crash-proofing + small robustness fixes

- pool.js: add pool.on('error') handler — an idle-client error (DB restart /
  .215 failover) previously crashed the process (no 'error' listener → throw)
- context tool: project a SAFE_COLUMNS allow-list for resources (never the
  monitoring/metadata JSON blobs); also add 'resource' to TABLE (was unhandled)
- applyPendingChange: guard the 'upsert' arm so a non-upsertable entity_type
  fails with a clear ValidationError instead of a bare TypeError

Tests: pool_error, context (resource case), pending_extended_actions (guard).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

tests/ai/agent/tools/context.test.js

@@ -23,4 +23,16 @@ describe('context tool', () => {
     const out = await contextTool.handler({}, { space_id: spaceId, view: null });
     expect(out.note).toMatch(/no specific entity/i);
   });
+
+  it('omits sensitive JSON blobs (monitoring/metadata) when grounding on a resource', async () => {
+    const { rows: [{ id: resId }] } = await pool.query(
+      `INSERT INTO resources(space_id, slug, name, runtime_type, monitoring, metadata)
+       VALUES($1,'db','DB','lxc','{"endpoint":"secret"}'::jsonb,'{"vault_path":"env:DB_PASS"}'::jsonb)
+       RETURNING id`, [spaceId]);
+    const out = await contextTool.handler({}, { space_id: spaceId, view: { entityType: 'resource', entityId: resId } });
+    expect(out.name).toBe('DB');
+    expect(out.status).toBe('unknown');
+    expect(out).not.toHaveProperty('monitoring');
+    expect(out).not.toHaveProperty('metadata');
+  });
 });