Raise the default model output limit to 16000 tokens

A full interview brief (RG-50.030.0002) needed about 6,000 output tokens,
so the old 6000 limit cut Claude's answer off and the run fell back to
the second model.

Claude-Session: https://claude.ai/code/session_01AteuhwysR4rHp4fqUteLAE

curatorbot/config.py

@@ -32,7 +32,7 @@ class Settings:
     # For the 'compatible' provider: an institution's own gateway or a local vLLM/Ollama server.
     llm_base_url: str = field(default_factory=lambda: os.getenv("CURATORBOT_LLM_BASE_URL", ""))
     llm_json_mode: bool = field(default_factory=lambda: _flag("CURATORBOT_LLM_JSON_MODE", True))
-    llm_max_tokens: int = field(default_factory=lambda: int(os.getenv("CURATORBOT_LLM_MAX_TOKENS", "6000")))
+    llm_max_tokens: int = field(default_factory=lambda: int(os.getenv("CURATORBOT_LLM_MAX_TOKENS", "16000")))
     ner_model: str = field(default_factory=lambda: os.getenv("CURATORBOT_NER_MODEL", "en_core_web_sm"))
     data_dir: Path = field(default_factory=lambda: Path(os.getenv("CURATORBOT_DATA_DIR", str(Path(__file__).resolve().parent.parent / "data" / "interviews"))))
     # Private Hugging Face dataset holding real briefs (e.g. ignatius/curatorbot-data).

docs/model-providers.md

@@ -17,7 +17,7 @@ Other settings:
 - `CURATORBOT_LLM_FALLBACK_MODEL`: tried once if the main model fails for a reason another model could fix (an outage, a rate limit, an unreadable answer). It is not tried for account problems such as a wrong key or no credit. Set it to an empty value to switch the fallback off. Briefs record which model actually answered.
 - `CURATORBOT_LLM_BASE_URL`: the address of a `compatible` endpoint, for example `http://localhost:11434/v1` for Ollama.
 - `CURATORBOT_LLM_JSON_MODE=0`: for local servers that reject the JSON-mode option.
-- `CURATORBOT_LLM_MAX_TOKENS`: raise it if long interviews return incomplete answers.
+- `CURATORBOT_LLM_MAX_TOKENS` (default 16000): the longest answer the model may write. A full interview's brief needs about 6,000; if an answer is cut off, the run falls back to the second model, which the brief records.
 
 `/api/health` shows the active provider, model and fallback.