feat(podcasts): feed poller + podcast_poll_interval_seconds setting (task 1.6)

One ticker re-reading its interval from the settings registry each
cycle (live tunable — the cache stays honest because writes go through
the settings API path, AGENTS.md gotcha #2). Per tick: list feeds
(first-admin identity, invariant 12), skip is_complete feeds (back
off; explicit refresh still works), poll each (conditional GET via
stored validators), and apply the subscriber auto-download policy
union after each successful fetch so new episodes queue without
waiting a tick.

Schema/registry: podcast_poll_interval_seconds (default 3600s, floor
300s, category podcasts) — seeded with ON CONFLICT DO NOTHING per the
idempotency rule.

internal/database/queries/queries.sql

@@ -4943,7 +4943,7 @@ WHERE source_feed_id = $1
 -- source filter keeps scan items out of podcast search.
 SELECT * FROM media_items
 WHERE source = 'podcast_episode'
-  AND ($1 = '' OR title ILIKE '%' || $1 || '%' OR description ILIKE '%' || $1 || '%')
+  AND (sqlc.arg('search_query')::text = '' OR title ILIKE '%' || sqlc.arg('search_query')::text || '%' OR description ILIKE '%' || sqlc.arg('search_query')::text || '%')
 ORDER BY feed_position ASC NULLS LAST, date_published ASC NULLS LAST
 LIMIT 50;
 
@@ -4957,9 +4957,43 @@ SELECT f.*,
        s.created_at AS subscribed_at
 FROM podcast_subscriptions s
 JOIN podcast_feeds f ON f.id = s.feed_id
-WHERE s.user_id = $1
-  AND ($2 = '' OR f.title ILIKE '%' || $2 || '%'
-       OR COALESCE(s.display_name, '') ILIKE '%' || $2 || '%'
-       OR COALESCE(f.host, '') ILIKE '%' || $2 || '%'
-       OR COALESCE(f.publisher, '') ILIKE '%' || $2 || '%')
+WHERE s.user_id = sqlc.arg('user_id')
+  AND (sqlc.arg('search_query')::text = '' OR f.title ILIKE '%' || sqlc.arg('search_query')::text || '%'
+       OR COALESCE(s.display_name, '') ILIKE '%' || sqlc.arg('search_query')::text || '%'
+       OR COALESCE(f.host, '') ILIKE '%' || sqlc.arg('search_query')::text || '%'
+       OR COALESCE(f.publisher, '') ILIKE '%' || sqlc.arg('search_query')::text || '%')
 ORDER BY COALESCE(s.display_name, f.title) ASC;
+
+-- name: GetNextQueuedEpisode :one
+-- Sequential download worker's claim source (single worker, day one).
+SELECT * FROM media_items
+WHERE source = 'podcast_episode' AND download_state = 'queued'
+ORDER BY updated_at ASC
+LIMIT 1;
+
+-- name: CountQueuedEpisodes :one
+SELECT COUNT(*) FROM media_items
+WHERE source = 'podcast_episode' AND download_state IN ('queued', 'downloading');
+
+-- name: CountPodcastEpisodesForFeed :one
+SELECT COUNT(*) FROM media_items
+WHERE source_feed_id = $1 AND source = 'podcast_episode';
+
+-- ============================================================================
+-- GENERIC reader_settings KV (podcast order override, plan D5: key
+-- podcast_order:<feed_id>). The pre-existing reader_settings queries are
+-- hard-keyed to 'reader_settings'; these two are the generic form.
+-- ============================================================================
+
+-- name: GetUserSettingByKey :one
+SELECT setting_value FROM reader_settings
+WHERE user_id = $1 AND setting_key = $2;
+
+-- name: UpsertUserSettingByKey :exec
+INSERT INTO reader_settings (user_id, setting_key, setting_value)
+VALUES ($1, $2, $3)
+ON CONFLICT (user_id, setting_key)
+DO UPDATE SET setting_value = EXCLUDED.setting_value, updated_at = NOW();
+
+-- name: DeleteUserSettingByKey :exec
+DELETE FROM reader_settings WHERE user_id = $1 AND setting_key = $2;

internal/database/settings_registry.go

@@ -82,6 +82,7 @@ var SettingDefaults = []SettingDefault{
 	{Key: "sync_queue_batch_size", Value: "50", Type: SettingTypeInt, Min: "1", Max: "10000", RequiresRestart: true, Category: "sync", Group: "Sync Queue", Description: "Maximum items processed per sync queue flush"},
 	{Key: "worker_pool_size", Value: "3", Type: SettingTypeInt, Min: "1", Max: "100", RequiresRestart: true, Category: "performance", Group: "Worker Pool", Description: "Number of background worker goroutines"},
 	{Key: "worker_queue_cap", Value: "100", Type: SettingTypeInt, Min: "1", Max: "10000", RequiresRestart: true, Category: "performance", Group: "Worker Pool", Description: "Background worker job queue capacity"},
+	{Key: "podcast_poll_interval_seconds", Value: "3600", Type: SettingTypeInt, Min: "300", Max: "604800", Category: "podcasts", Group: "Podcasts", Description: "How often to poll podcast feeds for new episodes (seconds; complete feeds back off)"},
 }
 
 // defaultBy indexes SettingDefaults by key for O(1) lookup.
@@ -188,6 +189,13 @@ func (r *SettingsRegistry) ScanPollInterval() time.Duration {
 	return time.Duration(r.getInt("scan_poll_interval_seconds")) * time.Second
 }
 
+// PodcastPollInterval is the feed-poller tick (live setting; the poller
+// re-reads it every cycle — writes go through the settings API path,
+// AGENTS.md gotcha #2).
+func (r *SettingsRegistry) PodcastPollInterval() time.Duration {
+	return time.Duration(r.getInt("podcast_poll_interval_seconds")) * time.Second
+}
+
 // AutoScanEnabled reports whether auto-scanning is on.
 func (r *SettingsRegistry) AutoScanEnabled() bool { return r.getBool("auto_scan_enabled") }