feat(openrouter): add live model sync

scripts/sync_openrouter_models.py

@@ -0,0 +1,155 @@
+#!/usr/bin/env python3
+
+import argparse
+import json
+import sys
+from pathlib import Path
+from urllib.error import HTTPError, URLError
+from urllib.request import Request, urlopen
+
+from providers.shared.temperature import TemperatureConstraint
+
+
+ROOT = Path(__file__).resolve().parents[1]
+DEFAULT_OUTPUT = ROOT / "conf" / "openrouter_models_live.json"
+OPENROUTER_MODELS_URL = "https://openrouter.ai/api/v1/models"
+
+
+def fetch_openrouter_models(url: str) -> dict:
+    request = Request(
+        url,
+        headers={
+            "Accept": "application/json",
+            "User-Agent": "pal-mcp-server/openrouter-model-sync",
+        },
+    )
+
+    with urlopen(request, timeout=30) as response:
+        charset = response.headers.get_content_charset("utf-8")
+        payload = response.read().decode(charset)
+
+    data = json.loads(payload)
+    if not isinstance(data, dict):
+        raise ValueError("OpenRouter models payload must be a JSON object")
+    return data
+
+
+def _supports_parameter(model_data: dict, parameter: str) -> bool:
+    supported = model_data.get("supported_parameters")
+    return isinstance(supported, list) and parameter in supported
+
+
+def _input_modalities(model_data: dict) -> list[str]:
+    architecture = model_data.get("architecture")
+    if not isinstance(architecture, dict):
+        return []
+
+    modalities = architecture.get("input_modalities")
+    if not isinstance(modalities, list):
+        return []
+
+    return [str(item) for item in modalities]
+
+
+def _infer_temperature_fields(model_name: str, model_data: dict) -> tuple[bool, str]:
+    if _supports_parameter(model_data, "temperature"):
+        return True, "range"
+
+    supports_temperature, _constraint, _reason = TemperatureConstraint.resolve_settings(model_name)
+    return supports_temperature, "fixed" if not supports_temperature else "range"
+
+
+def convert_model(model_data: dict) -> dict | None:
+    model_name = model_data.get("id")
+    if not isinstance(model_name, str) or not model_name:
+        return None
+
+    modalities = _input_modalities(model_data)
+    supports_temperature, temperature_constraint = _infer_temperature_fields(model_name, model_data)
+
+    context_window = model_data.get("context_length") or 0
+    top_provider = model_data.get("top_provider")
+    if not isinstance(top_provider, dict):
+        top_provider = {}
+
+    max_output_tokens = top_provider.get("max_completion_tokens") or 0
+
+    return {
+        "model_name": model_name,
+        "aliases": [],
+        "context_window": int(context_window) if context_window else 0,
+        "max_output_tokens": int(max_output_tokens) if max_output_tokens else 0,
+        "supports_extended_thinking": _supports_parameter(model_data, "reasoning")
+        or _supports_parameter(model_data, "include_reasoning"),
+        "supports_json_mode": _supports_parameter(model_data, "response_format")
+        or _supports_parameter(model_data, "structured_outputs"),
+        "supports_function_calling": _supports_parameter(model_data, "tools"),
+        "supports_images": "image" in modalities,
+        "max_image_size_mb": 20.0 if "image" in modalities else 0.0,
+        "supports_temperature": supports_temperature,
+        "temperature_constraint": temperature_constraint,
+        "description": model_data.get("description") or model_data.get("name") or "",
+        "intelligence_score": 10,
+        "allow_code_generation": False,
+    }
+
+
+def build_output_document(source: dict, source_url: str) -> dict:
+    models = []
+    for model_data in source.get("data", []):
+        if not isinstance(model_data, dict):
+            continue
+
+        converted = convert_model(model_data)
+        if converted:
+            models.append(converted)
+
+    models.sort(key=lambda item: item["model_name"])
+
+    return {
+        "_README": {
+            "description": "Generated baseline OpenRouter catalogue for PAL MCP Server.",
+            "source": source_url,
+            "usage": "Generated by scripts/sync_openrouter_models.py. Curated overrides belong in conf/openrouter_models.json.",
+            "field_notes": "Entries are conservative and intended as discovery data only. Curated manifest values override these at runtime.",
+        },
+        "models": models,
+    }
+
+
+def write_output(path: Path, document: dict) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    with path.open("w", encoding="utf-8", newline="\n") as handle:
+        json.dump(document, handle, indent=2, ensure_ascii=False)
+        handle.write("\n")
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description="Sync OpenRouter live model catalogue into PAL config.")
+    parser.add_argument("--url", default=OPENROUTER_MODELS_URL, help="OpenRouter models endpoint")
+    parser.add_argument(
+        "--output",
+        default=str(DEFAULT_OUTPUT),
+        help="Path to the generated live OpenRouter manifest",
+    )
+    return parser.parse_args()
+
+
+def main() -> int:
+    args = parse_args()
+    output_path = Path(args.output)
+
+    try:
+        source = fetch_openrouter_models(args.url)
+        document = build_output_document(source, args.url)
+        write_output(output_path, document)
+    except (HTTPError, URLError, TimeoutError, ValueError, json.JSONDecodeError) as exc:
+        print(f"Failed to sync OpenRouter models: {exc}", file=sys.stderr)
+        return 1
+
+    print(f"Wrote {len(document['models'])} OpenRouter models to {output_path}")
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())