refine: x-tweet vs x-article source_type, 500ms rate limit (Ganymede)

- Distinguish tweets (source_type: x-tweet, format: social-media) from
  articles (source_type: x-article, format: article) based on content
  length and article marker presence
- 500ms delay between fetch_from_url calls in research path
- Keep standalone sources pure (no Rio analysis — circular dependency)

Pentagon-Agent: Epimetheus <3D35839A-7722-4740-B93D-51157F7D5E70>

telegram/bot.py

@@ -332,7 +332,9 @@ async def handle_research(msg, query: str, user, silent: bool = False):
 
     # Fetch full content for top tweets (not just search snippets)
     from x_client import fetch_from_url
-    for tweet in tweets[:5]:  # Top 5 by engagement
+    for i, tweet in enumerate(tweets[:5]):  # Top 5 by engagement
+        if i > 0:
+            await asyncio.sleep(0.5)  # Ganymede: 500ms between calls, polite to Ben's API
         url = tweet.get("url", "")
         if url:
             try:
@@ -766,7 +768,8 @@ def _archive_standalone_source(url: str, content: str, user):
 
     Separate from the conversation archive — this is the actual article/tweet
     entering the extraction pipeline as a proper source, attributed to the
-    contributor who shared it.
+    contributor who shared it. Ganymede: keep pure (no Rio analysis), two
+    source_types (x-tweet vs x-article).
     """
     try:
         username = user.username if user else "anonymous"
@@ -778,6 +781,11 @@ def _archive_standalone_source(url: str, content: str, user):
         if author_match:
             author = f"@{author_match.group(1)}"
 
+        # Distinguish tweet vs article (Ganymede: different extraction behavior)
+        is_article = "--- Article Content ---" in content and len(content) > 1000
+        source_type = "x-article" if is_article else "x-tweet"
+        fmt = "article" if is_article else "social-media"
+
         slug = re.sub(r"[^a-z0-9]+", "-", f"{author}-{url.split('/')[-1][:30]}".lower()).strip("-")
         filename = f"{date_str}-tg-shared-{slug}.md"
         source_path = Path(ARCHIVE_DIR) / filename
@@ -788,20 +796,20 @@ def _archive_standalone_source(url: str, content: str, user):
 
         source_content = f"""---
 type: source
-source_type: x-article
+source_type: {source_type}
 title: "{author} — shared via Telegram by @{username}"
 author: "{author}"
 url: "{url}"
 date: {date_str}
 domain: internet-finance
-format: article
+format: {fmt}
 status: unprocessed
 proposed_by: "@{username}"
 contribution_type: source-submission
-tags: [telegram-shared, x-article]
+tags: [telegram-shared, {source_type}]
 ---
 
-# {author} — Article/Thread
+# {author} — {'Article' if is_article else 'Tweet/Thread'}
 
 Shared by @{username} via Telegram.
 Source URL: {url}