From 9149ee12bfbc1fefc15bc8e686b595fa1d509192 Mon Sep 17 00:00:00 2001
From: Nick Sweeting <git@sweeting.me>
Date: Wed, 10 Jun 2026 10:43:52 -0700
Subject: [PATCH] Reopen crawls before queuing discovered snapshots

---
 archivebox/services/runner.py                         | 11 +++++++++++
 ...st_api_v1_cli_workflow_add_search_update_remove.py |  4 ++--
 2 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/archivebox/services/runner.py b/archivebox/services/runner.py
index 4b8218ad..1010582f 100644
--- a/archivebox/services/runner.py
+++ b/archivebox/services/runner.py
@@ -771,6 +771,17 @@ class CrawlRunner:
         if not discovered_urls:
             return
 
+        if self.crawl.status == self.crawl.StatusChoices.SEALED:
+            # Snapshot completion projectors can observe the root snapshot seal
+            # before the runner has consumed parser urls.jsonl output. A sealed
+            # crawl must not block those freshly discovered child snapshots; the
+            # runner is still inside the same crawl lifecycle and will seal it
+            # again after the discovered queue is empty.
+            await sync_to_async(self.crawl.update_and_requeue, thread_sensitive=True)(
+                status=self.crawl.StatusChoices.STARTED,
+                retry_at=timezone.now(),
+            )
+
         parent_snapshot = await sync_to_async(
             lambda: Snapshot.objects.select_related("crawl", "crawl__created_by").filter(id=snapshot_payload["id"]).first(),
             thread_sensitive=True,
diff --git a/archivebox/tests/test_api_v1_cli_workflow_add_search_update_remove.py b/archivebox/tests/test_api_v1_cli_workflow_add_search_update_remove.py
index ca5c3b70..175e4dd9 100644
--- a/archivebox/tests/test_api_v1_cli_workflow_add_search_update_remove.py
+++ b/archivebox/tests/test_api_v1_cli_workflow_add_search_update_remove.py
@@ -42,7 +42,7 @@ def test_cli_api_add_search_update_remove_over_server(tmp_path):
                 "tag": "api-cli",
                 "depth": 0,
                 "parser": "url_list",
-                "plugins": "wget",
+                "plugins": "parse_txt_urls,wget",
                 "update": True,
                 "overwrite": False,
                 "index_only": True,
@@ -129,7 +129,7 @@ def test_cli_api_add_search_update_remove_over_server(tmp_path):
             crawl = (crawl_obj.max_depth, crawl_obj.tags_str, crawl_obj.config) if crawl_obj else None
 
         assert crawl is not None
-        assert crawl[0] == 0
+        assert crawl[0] == 1
         assert crawl[1] == "api-cli"
         assert crawl[2]["INDEX_ONLY"] is True