Run #3366

{
  "ok": true,
  "failed": [],
  "hallucinated": [],
  "context_errors": []
}

{
  "pool_status": "see_db"
}

Tu es un agent de monitoring du pool agencecom.

Lessons pertinentes:
LESSONS APPRISES PERTINENTES :
- L453-HUB: Health endpoints K8s : separer liveness no-DB / readiness DB-1s alert / startup cached
- L328-CROSS-S150-blocD2-deploy: Bloc D2 cache layer DEPLOYED Sprint 2 P2 — REUSE Redis 7.0.15 BSD existing Contabo (skip Valkey doublon) + smoke 5/5 PASS via /api/memory/cache/health endpoint dedie
- L-S199-LL2-V52-SPEC-REVIEW: L-S199-LL2 — Review V5.2 Monitoring spec (3 BLOCKERS + 15 medium + 5 low = 23 findings, GO sous condition)
- L-S199-CROSS-V52-SPEC-DRAFTED: L-S199-CROSS-V52-SPEC-DRAFTED — V5.2 Monitoring event-driven spec push DB
- L-S189-LL2-observateur-recap: L-S189-LL2 — Observateur PHASE 1 recap 13+ cycles SILENT + cron clôture

Donnees pool actuelles:
see_db

CONTRAINTE STRICTE -- Tables autorisees UNIQUEMENT :
- agencecom.agents (id, name, agent_level, status, current_session_id, error_count_last_hour, total_sessions, total_errors, total_cost_usd, quarantined_until, acquired_at, last_released_at)
- agencecom.agent_task_runs (id, agent_id, status, started_at, completed_at, error_message, llm_tokens_in, llm_tokens_out, cost_usd, validation_result)

Toute autre table (v_agents_critical, agent_pool, healthcheck_config, error_log, file_locks, zombie_agents, agent_audit, v_agent_timeout_health, v_agents_timeout_risk, agent_health_log, redis_health_ping, auto_quarantine, missions, health_checks, agent_health, internal) NEXISTE PAS -- ne reference jamais de table absente de cette liste.

Identifie les agents zombies (busy >1h sans completion via started_at + status), error_count_last_hour > 5, ou drift quarantine (quarantined_until > now).
Donne 3 recommandations concretes en markdown bullet points avec actions SQL utilisant SEULEMENT les 2 tables autorisees ci-dessus.

Voici 3 actions concrètes de surveillance du pool `agencecom` basées uniquement sur les 2 tables autorisées.

- **Libérer les agents zombies (busy > 1h sans completion)**  
  Requêtes pour tuer les sessions bloquées et remettre les agents à l’état `idle`.  
  ```sql
  -- Marquer les tâches zombies comme échouées
  UPDATE agencecom.agent_task_runs
  SET status = 'zombie_killed',
      completed_at = NOW(),
      error_message = 'Killed by monitor: zombie >1h'
  WHERE id IN (
      SELECT a.current_session_id
      FROM agencecom.agents a
      JOIN agencecom.agent_task_runs r ON r.id = a.current_session_id
      WHERE a.status = 'busy'
        AND r.started_at < NOW() - INTERVAL '1 hour'
        AND r.completed_at IS NULL
  );

  -- Libérer les agents correspondants
  UPDATE agencecom.agents
  SET status = 'idle',
      current_session_id = NULL,
      last_released_at = NOW()
  WHERE id IN (
      SELECT a.id
      FROM agencecom.agents a
      JOIN agencecom.agent_task_runs r ON r.id = a.current_session_id
      WHERE a.status = 'busy'
        AND r.started_at < NOW() - INTERVAL '1 hour'
        AND r.completed_at IS NULL
  );
  ```

- **Mettre en quarantaine les agents ayant plus de 5 erreurs/heure**  
  Protection automatique pour sortir du pool les agents instables.  
  ```sql
  UPDATE agencecom.agents
  SET quarantined_until = NOW() + INTERVAL '1 hour',
      status = CASE WHEN status != 'quarantined' THEN 'quarantined' ELSE status END
  WHERE error_count_last_hour > 5
    AND (quarantined_until IS NULL OR quarantined_until <= NOW());
  ```

- **Corriger la dérive de quarantaine (agents quarantined_until > NOW mais sains)**  
  Réintégrer les agents qui ne sont plus en erreur et n’ont pas de session zombie active.  
  ```sql
  UPDATE agencecom.agents
  SET quarantined_until = NULL,
      status = 'idle',
      last_released_at = NOW()
  WHERE quarantined_until > NOW()
    AND error_count_last_hour <= 5
    AND (current_session_id IS NULL
         OR NOT EXISTS (
             SELECT 1 FROM agencecom.agent_task_runs
             WHERE id = agents.current_session_id
               AND started_at < NOW() - INTERVAL '1 hour'
               AND completed_at IS NULL
         ));
  ```

Ces 3 recommandations couvrent :  
- l’élimination des agents bloqués depuis plus d’une heure,  
- la mise en quarantaine des agents à fort taux d’erreur,  
- la reprise des agents dont la quarantaine n’a plus de justification (dérive).