perf: Limit max_tokens to 64K for demo stability

PERFORMANCE OPTIMIZATION:

Token Limits Updated:
- Claude Sonnet 4.5: 200K → 64K (demo stability)
- GPT-4o: 16K (unchanged)
- Groq models: 32K (unchanged)
- Gemini 2.0 Flash: 65K → 64K
- System prompt: Updated to reflect 64K limit

RATIONALE:
- Prevents timeout issues in HF Spaces
- Maintains quality while ensuring responsiveness
- 64K sufficient for most demo use cases
- Production Ectus-R still supports 200K

Changes:
- app.py line 125: max_tokens=64000
- app.py line 173: max_output_tokens: 64000
- SYSTEM_PROMPT: Context window updated

🤖 Generated with Claude Code

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

app.py

@@ -94,7 +94,7 @@ OUTPUT FORMAT:
 3. Dockerfile (if deployment mentioned)
 4. Brief README with usage instructions
 
-Context window: 200,000 tokens output - you can generate comprehensive solutions.
+Context window: 64,000 tokens output (demo limit) - you can generate comprehensive solutions.
 
 Be complete and thorough. Focus on quality and production-readiness."""
 
@@ -122,7 +122,7 @@ def generate_code_with_model(prompt: str, model_name: str, temperature: float =
             client = anthropic.Anthropic(api_key=os.getenv(config["api_key_env"]))
             response = client.messages.create(
                 model=config["model"],
-                max_tokens=200000,
+                max_tokens=64000,  # Limited for demo stability
                 temperature=temperature,
                 system=SYSTEM_PROMPT,
                 messages=[{"role": "user", "content": prompt}]
@@ -143,7 +143,7 @@ def generate_code_with_model(prompt: str, model_name: str, temperature: float =
                     {"role": "user", "content": prompt}
                 ],
                 temperature=temperature,
-                max_tokens=16000  # GPT-4o max is 16K
+                max_tokens=16000  # GPT-4o limit
             )
             generated_code = response.choices[0].message.content
             input_tokens = response.usage.prompt_tokens
@@ -158,7 +158,7 @@ def generate_code_with_model(prompt: str, model_name: str, temperature: float =
                     {"role": "user", "content": prompt}
                 ],
                 temperature=temperature,
-                max_tokens=32000  # Groq supports up to 32K
+                max_tokens=32000  # Groq limit (kept at 32K)
             )
             generated_code = response.choices[0].message.content
             input_tokens = response.usage.prompt_tokens
@@ -170,7 +170,7 @@ def generate_code_with_model(prompt: str, model_name: str, temperature: float =
             model = genai.GenerativeModel(config["model"])
             response = model.generate_content(
                 f"{SYSTEM_PROMPT}\n\nUser request: {prompt}",
-                generation_config={"temperature": temperature, "max_output_tokens": 65536}  # Gemini 2.0 Flash supports up to 8K (65536 is max for SDK)
+                generation_config={"temperature": temperature, "max_output_tokens": 64000}  # Gemini 2.0 Flash supports up to 8K (65536 is max for SDK)
             )
             generated_code = response.text
             input_tokens = response.usage_metadata.prompt_token_count