From 4bef9d2d1776c8f0b4a56c64e86e5dacc111995c Mon Sep 17 00:00:00 2001
From: ChemaVX <jmivanez@gmail.com>
Date: Mon, 4 May 2026 13:24:25 +0000
Subject: [PATCH] feat: queries DDG generadas por Claude en lugar de plantillas
 hardcodeadas

---
 src/scraper/exhaustive.py | 43 ++++++++++++++++++++++++++++++++++++---
 1 file changed, 40 insertions(+), 3 deletions(-)

diff --git a/src/scraper/exhaustive.py b/src/scraper/exhaustive.py
index 4b676d1..74ff47d 100644
--- a/src/scraper/exhaustive.py
+++ b/src/scraper/exhaustive.py
@@ -164,9 +164,8 @@ class ExhaustiveScraper:
         ]
         await asyncio.gather(*tasks, return_exceptions=True)
 
-    async def _seed_duckduckgo(self):
-        """Multiple DDG queries — fresh DDGS() per query to avoid cascading ratelimits"""
-        queries = [
+    async def _generate_ddg_queries(self) -> list[str]:
+        fallback = [
             self.topic,
             f"{self.topic} history facts",
             f"{self.topic} evidence analysis",
@@ -176,6 +175,44 @@ class ExhaustiveScraper:
             f"{self.topic} documentary",
             f"{self.topic} research study",
         ]
+
+        if not settings.anthropic_api_key:
+            return fallback
+
+        try:
+            import anthropic
+            client = anthropic.AsyncAnthropic(api_key=settings.anthropic_api_key)
+            prompt = (
+                f'Generate exactly 8 DuckDuckGo search queries to research: "{self.topic}"\n\n'
+                f'Rules:\n'
+                f'- Each query must be specific and distinct — no generic templates\n'
+                f'- Cover different angles: facts, history, official sources, criticism, '
+                f'technical details, recent developments, expert opinions, primary sources\n'
+                f'- Use the most specific terminology for this topic\n'
+                f'- Include the topic language naturally (if topic is in Spanish, '
+                f'mix Spanish and English queries for broader coverage)\n'
+                f'- Output ONLY the 8 queries, one per line, no numbering, '
+                f'no explanations, no markdown\n'
+            )
+            msg = await client.messages.create(
+                model=settings.claude_model,
+                max_tokens=300,
+                messages=[{"role": "user", "content": prompt}]
+            )
+            raw = msg.content[0].text.strip()
+            queries = [q.strip() for q in raw.split('\n') if q.strip()]
+            if self.topic not in queries:
+                queries = [self.topic] + queries[:7]
+            queries = queries[:8]
+            logger.info("DDG queries generated by Claude", queries=queries)
+            return queries
+        except Exception as e:
+            logger.warning("Claude query generation failed, using fallback", error=str(e))
+            return fallback
+
+    async def _seed_duckduckgo(self):
+        """Multiple DDG queries — fresh DDGS() per query to avoid cascading ratelimits"""
+        queries = await self._generate_ddg_queries()
         for query in queries:
             if self._stop:
                 break