@amalgm/automations 0.2.1 → 0.2.2

package/supabase/migrations/20260830010000_durable_step_retries.sql

@@ -0,0 +1,332 @@
+-- One Supabase run row is the queue, lease, durable step journal, and history.
+-- The target machine is the only executor.
+
+ALTER TABLE public.amalgm_automation_runs
+  ADD COLUMN IF NOT EXISTS retry_at timestamptz;
+
+DROP FUNCTION IF EXISTS list_amalgm_pending_runs(uuid, text);
+DROP FUNCTION IF EXISTS mark_amalgm_run_sent(uuid, uuid, text, timestamptz);
+DROP FUNCTION IF EXISTS update_amalgm_run(uuid, uuid, text, jsonb);
+
+ALTER TABLE public.amalgm_automation_runs
+  DROP CONSTRAINT IF EXISTS amalgm_automation_run_retry_state;
+ALTER TABLE public.amalgm_automation_runs
+  ADD CONSTRAINT amalgm_automation_run_retry_state CHECK (
+    status = 'pending' OR retry_at IS NULL
+  );
+
+DROP INDEX IF EXISTS public.amalgm_automation_runs_claimable;
+CREATE INDEX amalgm_automation_runs_claimable
+  ON public.amalgm_automation_runs(user_id, target_id, status, retry_at, lease_expires_at, created_at, id)
+  WHERE status IN ('pending', 'sent', 'running');
+
+CREATE OR REPLACE FUNCTION list_amalgm_event_triggers(p_user_id uuid, p_target_id text)
+RETURNS TABLE (
+  user_id uuid,
+  automation_id text,
+  trigger_id text,
+  target_id text,
+  source text,
+  event text,
+  secret text
+)
+LANGUAGE sql
+SECURITY DEFINER
+SET search_path = public
+AS $$
+  SELECT t.user_id, t.automation_id, t.id, a.target_id, t.source, t.event, t.secret
+  FROM public.amalgm_automation_triggers t
+  JOIN public.amalgm_automations a
+    ON a.user_id = t.user_id AND a.id = t.automation_id
+  JOIN public.amalgm_automation_workflows w
+    ON w.user_id = a.user_id AND w.automation_id = a.id
+  WHERE t.kind = 'webhook' AND t.user_id = p_user_id
+    AND a.target_id = p_target_id AND t.enabled AND a.enabled
+    AND jsonb_typeof(w.compiled) = 'object'
+    AND w.compiled->>'version' = '1'
+    AND jsonb_typeof(w.compiled->'steps') = 'array'
+    AND jsonb_array_length(
+      CASE WHEN jsonb_typeof(w.compiled->'steps') = 'array'
+        THEN w.compiled->'steps' ELSE '[]'::jsonb END
+    ) BETWEEN 1 AND 100;
+$$;
+
+DROP FUNCTION IF EXISTS list_due_amalgm_crons(timestamptz);
+CREATE FUNCTION list_due_amalgm_crons(p_now timestamptz)
+RETURNS TABLE (
+  user_id uuid,
+  automation_id text,
+  trigger_id text,
+  target_id text,
+  cron text,
+  timezone text,
+  next_run_at timestamptz,
+  remaining_occurrences integer
+)
+LANGUAGE sql
+SECURITY DEFINER
+SET search_path = public
+AS $$
+  SELECT t.user_id, t.automation_id, t.id, a.target_id,
+    t.cron, t.timezone, t.next_run_at, t.remaining_occurrences
+  FROM public.amalgm_automation_triggers t
+  JOIN public.amalgm_automations a
+    ON a.user_id = t.user_id AND a.id = t.automation_id
+  JOIN public.amalgm_automation_workflows w
+    ON w.user_id = a.user_id AND w.automation_id = a.id
+  WHERE t.kind = 'schedule' AND t.next_run_at <= p_now
+    AND (t.remaining_occurrences IS NULL OR t.remaining_occurrences > 0)
+    AND t.enabled AND a.enabled
+    AND jsonb_typeof(w.compiled) = 'object'
+    AND w.compiled->>'version' = '1'
+    AND jsonb_typeof(w.compiled->'steps') = 'array'
+    AND jsonb_array_length(
+      CASE WHEN jsonb_typeof(w.compiled->'steps') = 'array'
+        THEN w.compiled->'steps' ELSE '[]'::jsonb END
+    ) BETWEEN 1 AND 100
+  ORDER BY t.next_run_at, t.user_id, t.automation_id, t.id;
+$$;
+
+-- Re-check the executable plan inside the same transaction that admits work.
+-- Discovery is advisory; this function is the authority at the race boundary.
+CREATE OR REPLACE FUNCTION enqueue_amalgm_run(
+  p_kind text,
+  p_user_id uuid,
+  p_automation_id text,
+  p_trigger_id text,
+  p_target_id text,
+  p_expected_next_run_at timestamptz,
+  p_next_run_at timestamptz,
+  p_input jsonb,
+  p_now timestamptz
+) RETURNS SETOF public.amalgm_automation_runs
+LANGUAGE plpgsql
+SECURITY DEFINER
+SET search_path = public
+AS $$
+DECLARE
+  v_inserted integer := 0;
+BEGIN
+  IF p_kind = 'cron' THEN
+    PERFORM 1
+    FROM public.amalgm_automation_triggers t
+    JOIN public.amalgm_automations a
+      ON a.user_id = t.user_id AND a.id = t.automation_id
+    WHERE t.user_id = p_user_id AND t.automation_id = p_automation_id
+      AND t.id = p_trigger_id AND a.target_id = p_target_id
+      AND t.kind = 'schedule' AND t.next_run_at = p_expected_next_run_at
+      AND (t.remaining_occurrences IS NULL OR t.remaining_occurrences > 0)
+      AND t.enabled AND a.enabled
+    FOR UPDATE OF t;
+  ELSIF p_kind = 'event' THEN
+    PERFORM 1
+    FROM public.amalgm_automation_triggers t
+    JOIN public.amalgm_automations a
+      ON a.user_id = t.user_id AND a.id = t.automation_id
+    WHERE t.user_id = p_user_id AND t.automation_id = p_automation_id
+      AND t.id = p_trigger_id AND a.target_id = p_target_id
+      AND t.kind = 'webhook' AND t.enabled AND a.enabled;
+  ELSE
+    RAISE EXCEPTION 'Invalid automation trigger kind';
+  END IF;
+  IF NOT FOUND THEN RETURN; END IF;
+
+  RETURN QUERY
+  INSERT INTO public.amalgm_automation_runs (
+    user_id, target_id, automation_id, trigger_id, workflow_id,
+    automation_payload, input, status, created_at
+  )
+  SELECT a.user_id, a.target_id, a.id, t.id, w.id,
+    jsonb_strip_nulls(jsonb_build_object(
+      'id', a.id, 'targetId', a.target_id, 'name', a.name,
+      'description', a.description, 'enabled', a.enabled,
+      'trigger', jsonb_strip_nulls(jsonb_build_object(
+        'id', t.id,
+        'kind', CASE WHEN t.kind = 'schedule' THEN 'cron' ELSE 'event' END,
+        'enabled', t.enabled, 'cron', t.cron, 'timezone', t.timezone,
+        'source', t.source, 'event', t.event
+      )),
+      'workflow', jsonb_strip_nulls(jsonb_build_object(
+        'id', w.id, 'name', w.name, 'script', w.script,
+        'compiled', w.compiled, 'allowlist', w.allowlist, 'limits', w.limits
+      ))
+    )),
+    p_input, 'pending', p_now
+  FROM public.amalgm_automations a
+  JOIN public.amalgm_automation_triggers t
+    ON t.user_id = a.user_id AND t.automation_id = a.id
+  JOIN public.amalgm_automation_workflows w
+    ON w.user_id = a.user_id AND w.automation_id = a.id
+  WHERE a.user_id = p_user_id AND a.id = p_automation_id AND t.id = p_trigger_id
+    AND a.target_id = p_target_id AND t.enabled AND a.enabled
+    AND jsonb_typeof(w.compiled) = 'object'
+    AND w.compiled->>'version' = '1'
+    AND jsonb_typeof(w.compiled->'steps') = 'array'
+    AND jsonb_array_length(
+      CASE WHEN jsonb_typeof(w.compiled->'steps') = 'array'
+        THEN w.compiled->'steps' ELSE '[]'::jsonb END
+    ) BETWEEN 1 AND 100
+  RETURNING *;
+  GET DIAGNOSTICS v_inserted = ROW_COUNT;
+
+  IF p_kind = 'cron' AND v_inserted = 1 THEN
+    UPDATE public.amalgm_automation_triggers SET
+      next_run_at = p_next_run_at,
+      remaining_occurrences = CASE WHEN remaining_occurrences IS NULL
+        THEN NULL ELSE remaining_occurrences - 1 END,
+      enabled = CASE WHEN remaining_occurrences = 1 THEN false ELSE enabled END,
+      updated_at = p_now
+    WHERE user_id = p_user_id AND automation_id = p_automation_id
+      AND id = p_trigger_id AND kind = 'schedule'
+      AND next_run_at = p_expected_next_run_at;
+  END IF;
+END;
+$$;
+
+CREATE OR REPLACE FUNCTION claim_amalgm_machine_runs(
+  p_user_id uuid,
+  p_target_id text,
+  p_limit integer,
+  p_lease_seconds integer
+) RETURNS SETOF public.amalgm_automation_runs
+LANGUAGE plpgsql
+SECURITY DEFINER
+SET search_path = public
+AS $$
+BEGIN
+  UPDATE public.amalgm_automation_runs SET
+    status = 'failed', error = COALESCE(error, 'Machine execution attempts exhausted'),
+    finished_at = now(), lease_token = NULL, lease_expires_at = NULL, retry_at = NULL
+  WHERE user_id = p_user_id AND target_id = p_target_id AND attempts >= 5
+    AND (
+      (status = 'pending' AND COALESCE(retry_at, '-infinity'::timestamptz) <= now())
+      OR (status IN ('sent', 'running') AND COALESCE(lease_expires_at, '-infinity'::timestamptz) <= now())
+    );
+
+  RETURN QUERY
+  WITH candidates AS (
+    SELECT id
+    FROM public.amalgm_automation_runs
+    WHERE user_id = p_user_id AND target_id = p_target_id AND attempts < 5
+      AND (
+        (status = 'pending' AND COALESCE(retry_at, '-infinity'::timestamptz) <= now())
+        OR (status IN ('sent', 'running') AND COALESCE(lease_expires_at, '-infinity'::timestamptz) <= now())
+      )
+    ORDER BY created_at, id
+    FOR UPDATE SKIP LOCKED
+    LIMIT LEAST(GREATEST(p_limit, 1), 20)
+  )
+  UPDATE public.amalgm_automation_runs r SET
+    status = 'sent', sent_at = COALESCE(sent_at, now()),
+    lease_token = gen_random_uuid(),
+    lease_expires_at = now() + make_interval(secs => LEAST(GREATEST(p_lease_seconds, 30), 900)),
+    retry_at = NULL,
+    attempts = attempts + 1
+  FROM candidates
+  WHERE r.id = candidates.id
+  RETURNING r.*;
+END;
+$$;
+
+CREATE OR REPLACE FUNCTION update_amalgm_claimed_run(
+  p_user_id uuid,
+  p_target_id text,
+  p_run_id uuid,
+  p_lease_token uuid,
+  p_update jsonb,
+  p_now timestamptz
+) RETURNS SETOF public.amalgm_automation_runs
+LANGUAGE plpgsql
+SECURITY DEFINER
+SET search_path = public
+AS $$
+DECLARE
+  v_status text := p_update->>'status';
+BEGIN
+  IF v_status NOT IN ('pending', 'running', 'completed', 'failed') THEN
+    RAISE EXCEPTION 'Invalid run status';
+  END IF;
+  IF p_update ? 'retryAt' AND v_status <> 'pending' THEN
+    RAISE EXCEPTION 'Only a pending run may have retryAt';
+  END IF;
+
+  RETURN QUERY
+  UPDATE public.amalgm_automation_runs SET
+    status = CASE WHEN v_status = 'pending' AND attempts >= 5 THEN 'failed' ELSE v_status END,
+    started_at = CASE WHEN v_status = 'running' THEN COALESCE(started_at, p_now) ELSE started_at END,
+    finished_at = CASE
+      WHEN v_status IN ('completed', 'failed') OR (v_status = 'pending' AND attempts >= 5) THEN p_now
+      WHEN v_status = 'pending' THEN NULL
+      ELSE finished_at
+    END,
+    output = CASE WHEN p_update ? 'output' THEN p_update->'output' ELSE output END,
+    error = CASE
+      WHEN p_update ? 'error' THEN left(p_update->>'error', 1000)
+      WHEN v_status = 'running' THEN NULL
+      ELSE error
+    END,
+    retry_at = CASE
+      WHEN v_status = 'pending' AND attempts < 5 AND p_update ? 'retryAt'
+        THEN (p_update->>'retryAt')::timestamptz
+      ELSE NULL
+    END,
+    lease_token = CASE WHEN v_status = 'running' THEN lease_token ELSE NULL END,
+    lease_expires_at = CASE WHEN v_status = 'running' THEN p_now + interval '60 seconds' ELSE NULL END
+  WHERE id = p_run_id AND user_id = p_user_id AND target_id = p_target_id
+    AND status IN ('sent', 'running') AND lease_token = p_lease_token
+    AND lease_expires_at > p_now
+  RETURNING *;
+END;
+$$;
+
+CREATE OR REPLACE FUNCTION list_amalgm_runs(
+  p_user_id uuid,
+  p_automation_id text,
+  p_status text,
+  p_limit integer,
+  p_offset integer
+)
+RETURNS jsonb
+LANGUAGE sql
+SECURITY DEFINER
+SET search_path = public
+AS $$
+  WITH filtered AS (
+    SELECT id, automation_id, trigger_id, workflow_id, target_id, status, attempts,
+      retry_at, input, created_at, sent_at, started_at, finished_at, output, error
+    FROM public.amalgm_automation_runs
+    WHERE user_id = p_user_id AND automation_id = p_automation_id
+      AND (p_status IS NULL OR status = p_status)
+  ), page_rows AS (
+    SELECT * FROM filtered ORDER BY created_at DESC, id DESC LIMIT p_limit OFFSET p_offset
+  )
+  SELECT jsonb_build_object(
+    'items', COALESCE((SELECT jsonb_agg(to_jsonb(page_rows)) FROM page_rows), '[]'::jsonb),
+    'total', (SELECT count(*) FROM filtered)
+  );
+$$;
+
+CREATE OR REPLACE FUNCTION get_amalgm_run(p_user_id uuid, p_automation_id text, p_run_id text)
+RETURNS jsonb
+LANGUAGE sql
+SECURITY DEFINER
+SET search_path = public
+AS $$
+  SELECT to_jsonb(r) - 'user_id' - 'automation_payload' - 'lease_token' - 'lease_expires_at'
+  FROM public.amalgm_automation_runs r
+  WHERE r.user_id = p_user_id AND r.automation_id = p_automation_id AND r.id::text = p_run_id;
+$$;
+
+REVOKE ALL ON FUNCTION claim_amalgm_machine_runs(uuid, text, integer, integer) FROM PUBLIC;
+REVOKE ALL ON FUNCTION update_amalgm_claimed_run(uuid, text, uuid, uuid, jsonb, timestamptz) FROM PUBLIC;
+REVOKE ALL ON FUNCTION list_due_amalgm_crons(timestamptz) FROM PUBLIC;
+REVOKE ALL ON FUNCTION list_amalgm_event_triggers(uuid, text) FROM PUBLIC;
+REVOKE ALL ON FUNCTION list_amalgm_runs(uuid, text, text, integer, integer) FROM PUBLIC;
+REVOKE ALL ON FUNCTION get_amalgm_run(uuid, text, text) FROM PUBLIC;
+
+GRANT EXECUTE ON FUNCTION claim_amalgm_machine_runs(uuid, text, integer, integer) TO service_role;
+GRANT EXECUTE ON FUNCTION update_amalgm_claimed_run(uuid, text, uuid, uuid, jsonb, timestamptz) TO service_role;
+GRANT EXECUTE ON FUNCTION list_due_amalgm_crons(timestamptz) TO service_role;
+GRANT EXECUTE ON FUNCTION list_amalgm_event_triggers(uuid, text) TO service_role;
+GRANT EXECUTE ON FUNCTION list_amalgm_runs(uuid, text, text, integer, integer) TO service_role;
+GRANT EXECUTE ON FUNCTION get_amalgm_run(uuid, text, text) TO service_role;