feat: add EN/ES/RO language selector for AI pitch generation

- db.py: add `language` column to ai_queue; migration; queue_ai() accepts
  language param and re-queues with ON CONFLICT UPDATE so changing language works
- main.py: batch and single assess endpoints accept `language` from request body
- enricher.py: ai_worker_loop reads language column, passes to _assess_one()
- replicate_ai.py: assess_domain() and _build_prompt() accept language param;
  OUTPUT LANGUAGE section injected into prompt so Gemini writes pitch/email in
  the requested language (EN/ES/RO)
- index.html: flag dropdown (🇪🇸/🇬🇧/🇷🇴) next to AI Assess button; aiLang
  state default ES; language sent in all batch assessment requests

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

app/replicate_ai.py

@@ -54,7 +54,7 @@ async def _ddg_search(query: str) -> str:
         return ""
 
 
-def _build_prompt(a: dict, search_results: str = "") -> str:
+def _build_prompt(a: dict, search_results: str = "", language: str = "ES") -> str:
     contacts_block = []
     if a.get("emails"):      contacts_block.append(f"  Emails:    {', '.join(a['emails'][:3])}")
     if a.get("phones"):      contacts_block.append(f"  Phones:    {', '.join(a['phones'][:3])}")
@@ -141,6 +141,10 @@ Profiles found on site: {social_str}
 === WEB SEARCH RESULTS (use to find contacts, verify business identity) ===
 {(search_results or "No results.")[:600]}
 
+=== OUTPUT LANGUAGE ===
+Write pitch_angle, outreach_email, email_subject, and all human-readable text fields in: {language}
+(EN = English | ES = Spanish | RO = Romanian)
+
 === WHO WE ARE ===
 We are a full-service digital agency. We handle EVERYTHING web-related for SMEs:
 new website builds, redesigns, landing pages, e-commerce, CMS migrations, speed
@@ -246,7 +250,7 @@ def _parse_output(raw: str) -> dict:
     }
 
 
-async def assess_domain(analysis: dict) -> dict:
+async def assess_domain(analysis: dict, language: str = "ES") -> dict:
     """Call Gemini with the full site analysis. Returns parsed assessment."""
     async with _sem():
         # Build search query from domain / page title for contact lookup
@@ -259,7 +263,7 @@ async def assess_domain(analysis: dict) -> dict:
 
         payload = {
             "input": {
-                "prompt": _build_prompt(analysis, search_results),
+                "prompt": _build_prompt(analysis, search_results, language),
                 "images":  [],
                 "videos":  [],
                 "top_p":   0.9,