Run #3392

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

{}

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:
[missing:pool_status]

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.

- **Recommandation 1 : Nettoyage des agents zombies (statut `busy` depuis >1h sans complétion)**  
  *Détection* :  
  ```sql
  SELECT a.id, a.name, r.id AS run_id, r.started_at
  FROM agencecom.agents a
  INNER JOIN agencecom.agent_task_runs r ON a.id = r.agent_id
  WHERE a.status = 'busy'
    AND r.completed_at IS NULL
    AND r.started_at < NOW() - INTERVAL '1 hour';
  ```
  *Action corrective* :  
  ```sql
  -- Libérer l'agent
  UPDATE agencecom.agents
  SET status = 'available', current_session_id = NULL
  WHERE id IN (
    SELECT a.id
    FROM agencecom.agents a
    INNER JOIN agencecom.agent_task_runs r ON a.id = r.agent_id
    WHERE a.status = 'busy'
      AND r.completed_at IS NULL
      AND r.started_at < NOW() - INTERVAL '1 hour'
  );

  -- Clore la tâche zombie en erreur
  UPDATE agencecom.agent_task_runs
  SET completed_at = NOW(),
      status = 'failed',
      error_message = 'Zombie timeout (tâche >1h sans complétion)'
  WHERE id IN (
    SELECT r.id
    FROM agencecom.agent_task_runs r
    JOIN agencecom.agents a ON a.id = r.agent_id
    WHERE a.status = 'busy'  -- l’agent était busy lors de la détection
      AND r.completed_at IS NULL
      AND r.started_at < NOW() - INTERVAL '1 hour'
  );
  ```

- **Recommandation 2 : Mise en quarantaine des agents à taux d’erreur élevé (>5 en une heure)**  
  *Détection* :  
  ```sql
  SELECT id, name, error_count_last_hour, status, quarantined_until
  FROM agencecom.agents
  WHERE error_count_last_hour > 5
    AND (quarantined_until IS NULL OR quarantined_until < NOW());
  ```
  *Action corrective* :  
  ```sql
  UPDATE agencecom.agents
  SET quarantined_until = NOW() + INTERVAL '30 minutes',
      status = 'available',
      current_session_id = NULL
  WHERE error_count_last_hour > 5
    AND (quarantined_until IS NULL OR quarantined_until < NOW());
  ```

- **Recommandation 3 : Correction de la dérive de quarantaine (agents avec quarantined_until > now mais encore occupés)**  
  *Détection de l’incohérence* :  
  ```sql
  SELECT id, name, status, quarantined_until, current_session_id
  FROM agencecom.agents
  WHERE quarantined_until > NOW() AND status != 'available';
  ```
  *Action corrective* :  
  ```sql
  -- Libérer l’agent et forcer le retour à available
  UPDATE agencecom.agents
  SET status = 'available', current_session_id = NULL
  WHERE quarantined_until > NOW() AND status != 'available';

  -- Si une tâche active est liée, la clore comme abandonnée
  UPDATE agencecom.agent_task_runs
  SET completed_at = NOW(),
      status = 'failed',
      error_message = 'Quarantine drift : tâche annulée (agent en quarantaine)'
  WHERE agent_id IN (
    SELECT id
    FROM agencecom.agents
    WHERE quarantined_until > NOW() AND status != 'available'
  )
    AND completed_at IS NULL;
  ```