diff --git a/.gitignore b/.gitignore
index 33b839a3bc..3b668c7bc9 100644
--- a/.gitignore
+++ b/.gitignore
@@ -37,6 +37,9 @@ crates/programs/aichatbot-sidecar/_ignored_schema.rs
 # AI Chat sidecar's SQlite database shouldn't be checked in
 crates/programs/aichatbot-sidecar/database.db
 
+# Runtime data for the AI chatbot
+runtime_data/
+
 # Old Python virtual environments
 models/Wav2Lip/python/
 models/tacotron2/python/
diff --git a/crates/programs/aichatbot-sidecar/_sql/migrations/2023-02-05-064307_web_scraping_targets/up.sql b/crates/programs/aichatbot-sidecar/_sql/migrations/2023-02-05-064307_web_scraping_targets/up.sql
index 7d34e262cb..b49b720246 100644
--- a/crates/programs/aichatbot-sidecar/_sql/migrations/2023-02-05-064307_web_scraping_targets/up.sql
+++ b/crates/programs/aichatbot-sidecar/_sql/migrations/2023-02-05-064307_web_scraping_targets/up.sql
@@ -14,6 +14,14 @@ CREATE TABLE web_scraping_targets (
   -- We'll need to figure this out for each site.
   canonical_url TEXT NOT NULL UNIQUE,
 
+  -- A full image will only be set at this stage if the RSS feed or HTML specifies one.
+  -- This can be useful downstream during the actual scraping.
+  maybe_article_full_image_url TEXT,
+
+  -- A thumbnail will only be set at this stage if the RSS feed or HTML specifies one.
+  -- This can be useful downstream during the actual scraping.
+  maybe_article_thumbnail_image_url TEXT,
+
   -- Scraping status: "new", "failed", "permanently_failed", "success"
   scraping_status TEXT DEFAULT "new" NOT NULL,
   scrape_attempts INT NOT NULL,
diff --git a/data/.gitkeep b/runtime_data/.gitkeep
similarity index 100%
rename from data/.gitkeep
rename to runtime_data/.gitkeep