rasa-pro 3.14.0a1__py3-none-any.whl → 3.14.0a2__py3-none-any.whl

rasa/builder/copilot/copilot.py

@@ -55,29 +55,7 @@ class Copilot:
     @asynccontextmanager
     async def _get_client(self) -> AsyncGenerator[openai.AsyncOpenAI, None]:
         """Create a fresh OpenAI client, yield it, and always close it."""
-        kwargs: Dict[str, Any] = {"timeout": config.OPENAI_TIMEOUT}
-        if config.HELLO_LLM_PROXY_BASE_URL:
-            structlogger.debug(
-                "copilot.using_llm_proxy", base_url=config.HELLO_LLM_PROXY_BASE_URL
-            )
-            if not config.RASA_PRO_LICENSE:
-                structlogger.error(
-                    "copilot.proxy_missing_license",
-                    event_info=(
-                        "HELLO_LLM_PROXY_BASE_URL is set "
-                        "but RASA_PRO_LICENSE is missing."
-                    ),
-                )
-                raise CopilotStreamError(
-                    "HELLO_LLM_PROXY_BASE_URL is set but RASA_PRO_LICENSE is missing. "
-                    "Provide a valid license token for proxy authentication."
-                )
-
-            kwargs["base_url"] = config.HELLO_LLM_PROXY_BASE_URL
-            kwargs["api_key"] = config.RASA_PRO_LICENSE
-
-        client = openai.AsyncOpenAI(**kwargs)
-
+        client = openai.AsyncOpenAI(timeout=config.OPENAI_TIMEOUT)
         try:
             yield client
         except Exception as e:

rasa/builder/download.py

@@ -20,11 +20,6 @@ def _get_python_version_content() -> str:
     return f"{sys.version_info.major}.{sys.version_info.minor}\n"
 
 
-def _get_rasa_pro_minor_version() -> str:
-    """Get the minor version of Rasa Pro."""
-    return ".".join(rasa.__version__.split(".")[:2])
-
-
 def _get_pyproject_toml_content(project_id: str) -> str:
     """Generate pyproject.toml file content."""
     return dedent(
@@ -34,7 +29,7 @@ def _get_pyproject_toml_content(project_id: str) -> str:
         version = "0.1.0"
         description = "Add your description for your Rasa bot here"
         readme = "README.md"
-        dependencies = ["rasa-pro>={_get_rasa_pro_minor_version()}"]
+        dependencies = ["rasa-pro>={rasa.__version__}"]
         requires-python = ">={sys.version_info.major}.{sys.version_info.minor}"
         """
     )

rasa/builder/main.py

@@ -22,6 +22,9 @@ from rasa.builder.logging_utils import (
     log_request_start,
 )
 from rasa.builder.service import bp, setup_project_generator
+from rasa.builder.template_cache import (
+    background_download_template_caches,
+)
 from rasa.builder.training_service import try_load_existing_agent
 from rasa.core.channels.studio_chat import StudioChatInput
 from rasa.server import configure_cors
@@ -146,9 +149,46 @@ def create_app(project_folder: str) -> Sanic:
         except Exception as e:
             structlogger.warning("Failed to load agent on server startup", error=str(e))
 
+    if config.HELLO_RASA_PROJECT_ID and app.ctx.project_generator.is_empty():
+        app.register_listener(background_download_template_caches, "after_server_start")
+    else:
+        structlogger.debug(
+            "builder.main.background_cache_download.disabled",
+            event_info=(
+                "No hello rasa project id set; skipping background cache download"
+            ),
+        )
+
     return app
 
 
+def _apply_llm_overrides_from_builder_env() -> None:
+    # Prefer a dedicated builder key, fall back to license if you proxy with it
+    if not config.HELLO_LLM_PROXY_BASE_URL:
+        return
+
+    structlogger.debug(
+        "builder.main.using_llm_proxy", base_url=config.HELLO_LLM_PROXY_BASE_URL
+    )
+
+    if not config.RASA_PRO_LICENSE:
+        structlogger.error(
+            "copilot.proxy_missing_license",
+            event_info=(
+                "HELLO_LLM_PROXY_BASE_URL is set but RASA_PRO_LICENSE is missing."
+            ),
+        )
+        return
+
+    if not os.getenv("OPENAI_API_BASE") and not os.getenv("OPENAI_API_KEY"):
+        base_url = config.HELLO_LLM_PROXY_BASE_URL.rstrip("/")
+        # needed for litellm client
+        os.environ["OPENAI_API_BASE"] = base_url
+        # needed for openai async client
+        os.environ["OPENAI_BASE_URL"] = base_url
+        os.environ["OPENAI_API_KEY"] = config.RASA_PRO_LICENSE
+
+
 def main(project_folder: Optional[str] = None) -> None:
     """Main entry point."""
     try:
@@ -159,6 +199,9 @@ def main(project_folder: Optional[str] = None) -> None:
         rasa.telemetry.initialize_telemetry()
         rasa.telemetry.initialize_error_reporting(private_mode=False)
 
+        # TODO: don't do this when running locally
+        _apply_llm_overrides_from_builder_env()
+
         # working directory needs to be the project folder, e.g.
         # for relative paths (./docs) in a projects config to work
         if not project_folder:

rasa/builder/project_generator.py

@@ -3,23 +3,18 @@
 import json
 import os
 import shutil
-import tarfile
-import tempfile
 from pathlib import Path
 from textwrap import dedent
-from typing import Any, Dict, Generator, List, Optional
+from typing import Any, Dict, List, Optional
 
-import aiofiles
-import aiohttp
 import structlog
 
-import rasa.version
 from rasa.builder import config
 from rasa.builder.exceptions import ProjectGenerationError, ValidationError
 from rasa.builder.llm_service import get_skill_generation_messages, llm_service
-from rasa.builder.logging_utils import capture_exception_with_context
 from rasa.builder.models import BotFiles
 from rasa.builder.project_info import ProjectInfo, ensure_first_used, load_project_info
+from rasa.builder.template_cache import copy_cache_for_template_if_available
 from rasa.builder.training_service import TrainingInput
 from rasa.builder.validation_service import validate_project
 from rasa.cli.scaffold import ProjectTemplateName, create_initial_project
@@ -48,12 +43,25 @@ class ProjectGenerator:
         """Get the project info."""
         return load_project_info(self.project_folder)
 
+    def is_empty(self) -> bool:
+        """Check if the project folder is empty.
+
+        Excluding hidden paths.
+        """
+        return not any(
+            file.is_file()
+            for file in self.project_folder.iterdir()
+            if not file.name.startswith(".")
+        )
+
     async def init_from_template(self, template: ProjectTemplateName) -> None:
         """Create the initial project files."""
         self.cleanup()
         create_initial_project(self.project_folder.as_posix(), template)
-        await download_cache_for_template(template, self.project_folder.as_posix())
-        # needs to happen after caching, as we download .rasa and that would
+        # If a local cache for this template exists, copy it into the project.
+        # We no longer download here to avoid blocking project creation.
+        copy_cache_for_template_if_available(template, self.project_folder)
+        # needs to happen after caching, as we download/copy .rasa and that would
         # overwrite the project info file in .rasa
         ensure_first_used(self.project_folder)
 
@@ -341,122 +349,3 @@ class ProjectGenerator:
                     error=str(e),
                     file_path=file_path,
                 )
-
-
-CACHE_BUCKET_URL = "https://trained-templates.s3.us-east-1.amazonaws.com"
-
-
-def _safe_tar_members(
-    tar: tarfile.TarFile, destination_directory: Path
-) -> Generator[tarfile.TarInfo, None, None]:
-    """Yield safe members for extraction to prevent path traversal and links.
-
-    Args:
-        tar: Open tar file handle
-        destination_directory: Directory to which files will be extracted
-
-    Yields:
-        Members that are safe to extract within destination_directory
-    """
-    base_path = destination_directory.resolve()
-
-    for member in tar.getmembers():
-        name = member.name
-        # Skip empty names and absolute paths
-        if not name or name.startswith("/") or name.startswith("\\"):
-            continue
-
-        # Disallow symlinks and hardlinks
-        if member.issym() or member.islnk():
-            continue
-
-        # Compute the final path and ensure it's within base_path
-        target_path = (base_path / name).resolve()
-        try:
-            target_path.relative_to(base_path)
-        except ValueError:
-            # Member would escape the destination directory
-            continue
-
-        yield member
-
-
-async def download_cache_for_template(
-    template: ProjectTemplateName, project_folder: str
-) -> None:
-    # get a temp path for the cache file download
-    temporary_cache_file = tempfile.NamedTemporaryFile(suffix=".tar.gz", delete=False)
-
-    try:
-        url = f"{CACHE_BUCKET_URL}/{rasa.version.__version__}-{template.value}.tar.gz"
-        async with aiohttp.ClientSession() as session:
-            async with session.get(url) as response:
-                response.raise_for_status()
-                async with aiofiles.open(temporary_cache_file.name, "wb") as f:
-                    async for chunk in response.content.iter_chunked(1024 * 1024):
-                        await f.write(chunk)
-
-        # extract the cache to the project folder using safe member filtering
-        with tarfile.open(temporary_cache_file.name, "r:gz") as tar:
-            destination = Path(project_folder)
-            destination.mkdir(parents=True, exist_ok=True)
-            tar.extractall(
-                path=destination,
-                members=_safe_tar_members(tar, destination),
-            )
-
-        structlogger.info(
-            "project_generator.download_cache_for_template.success",
-            template=template,
-            event_info=(
-                f"Downloaded cache for template, extracted to {project_folder}."
-            ),
-        )
-    except aiohttp.ClientResponseError as e:
-        if e.status == 403:
-            structlogger.debug(
-                "project_generator.download_cache_for_template.no_cache_found",
-                template=template,
-                event_info=("No cache found for template, continuing without it."),
-            )
-        else:
-            structlogger.debug(
-                "project_generator.download_cache_for_template.response_error",
-                error=str(e),
-                status=e.status,
-                template=template,
-                event_info=(
-                    "Failed to download cache for template, continuing without it."
-                ),
-            )
-            capture_exception_with_context(
-                e,
-                "project_generator.download_cache_for_template.response_error",
-                tags={"template": template.value, "status": str(e.status)},
-            )
-    except Exception as exc:
-        structlogger.debug(
-            "project_generator.download_cache_for_template.unexpected_error",
-            error=str(exc),
-            template=template,
-            event_info=(
-                "Unexpected error when downloading cache for template, "
-                "continuing without it."
-            ),
-        )
-        capture_exception_with_context(
-            exc,
-            "project_generator.download_cache_for_template.unexpected_error",
-            tags={"template": template.value},
-        )
-    finally:
-        # Clean up the temporary file
-        try:
-            Path(temporary_cache_file.name).unlink(missing_ok=True)
-        except Exception as exc:
-            structlogger.debug(
-                "project_generator.download_cache_for_template.cleanup_error",
-                error=str(exc),
-                template=template,
-                event_info=("Failed to cleanup cache for template, ignoring."),
-            )

rasa/builder/template_cache.py

@@ -0,0 +1,244 @@
+import asyncio
+import os
+import shutil
+import tarfile
+import tempfile
+from pathlib import Path
+from typing import Generator
+
+import aiofiles
+import aiohttp
+import structlog
+from sanic import Sanic
+
+import rasa.version
+from rasa.builder.logging_utils import capture_exception_with_context
+from rasa.cli.scaffold import ProjectTemplateName
+
+structlogger = structlog.get_logger()
+
+CACHE_BUCKET_URL = "https://trained-templates.s3.us-east-1.amazonaws.com"
+
+# Root directory for storing downloaded template caches on disk.
+_CACHE_ROOT_DIR = Path(
+    os.getenv(
+        "RASA_TEMPLATE_CACHE_DIR",
+        Path.home().joinpath(".rasa", "template-cache").as_posix(),
+    )
+)
+
+
+def _template_cache_dir(template: ProjectTemplateName) -> Path:
+    """Return the local cache directory for a given template and version."""
+    return _CACHE_ROOT_DIR / rasa.version.__version__ / template.value
+
+
+def _cache_root_dir() -> Path:
+    return Path(
+        os.getenv(
+            "RASA_TEMPLATE_CACHE_DIR",
+            Path.home().joinpath(".rasa", "template-cache").as_posix(),
+        )
+    )
+
+
+def _safe_tar_members(
+    tar: tarfile.TarFile, destination_directory: Path
+) -> Generator[tarfile.TarInfo, None, None]:
+    """Yield safe members for extraction to prevent path traversal and links.
+
+    Args:
+        tar: Open tar file handle
+        destination_directory: Directory to which files will be extracted
+
+    Yields:
+        Members that are safe to extract within destination_directory
+    """
+    base_path = destination_directory.resolve()
+
+    for member in tar.getmembers():
+        name = member.name
+        # Skip empty names and absolute paths
+        if not name or name.startswith("/") or name.startswith("\\"):
+            continue
+
+        # Disallow symlinks and hardlinks
+        if member.issym() or member.islnk():
+            continue
+
+        # Compute the final path and ensure it's within base_path
+        target_path = (base_path / name).resolve()
+        try:
+            target_path.relative_to(base_path)
+        except ValueError:
+            # Member would escape the destination directory
+            continue
+
+        yield member
+
+
+def _copytree(src: Path, dst: Path) -> None:
+    """Copy directory tree from src to dst, merging into dst.
+
+    Existing files are overwritten. Hidden files and directories are included, as
+    caches can contain `.rasa` metadata that should be applied before calling
+    `ensure_first_used`.
+    """
+    for root, dirs, files in os.walk(src):
+        rel_path = Path(root).relative_to(src)
+        target_dir = dst / rel_path
+        target_dir.mkdir(parents=True, exist_ok=True)
+        for filename in files:
+            src_file = Path(root) / filename
+            dst_file = target_dir / filename
+            shutil.copy2(src_file, dst_file)
+
+
+async def download_cache_for_template(
+    template: ProjectTemplateName, target_dir: str
+) -> None:
+    # get a temp path for the cache file download
+    temporary_cache_file = tempfile.NamedTemporaryFile(suffix=".tar.gz", delete=False)
+
+    try:
+        url = f"{CACHE_BUCKET_URL}/{rasa.version.__version__}-{template.value}.tar.gz"
+        async with aiohttp.ClientSession() as session:
+            async with session.get(url) as response:
+                response.raise_for_status()
+                async with aiofiles.open(temporary_cache_file.name, "wb") as f:
+                    async for chunk in response.content.iter_chunked(1024 * 1024):
+                        await f.write(chunk)
+
+        # extract the cache to the project folder using safe member filtering
+        with tarfile.open(temporary_cache_file.name, "r:gz") as tar:
+            destination = Path(target_dir)
+            destination.mkdir(parents=True, exist_ok=True)
+            tar.extractall(
+                path=destination,
+                members=_safe_tar_members(tar, destination),
+            )
+
+        structlogger.info(
+            "project_generator.download_cache_for_template.success",
+            template=template,
+            event_info=(
+                "Downloaded cache for template, extracted to target directory."
+            ),
+            target_dir=target_dir,
+        )
+    except aiohttp.ClientResponseError as e:
+        if e.status == 403:
+            structlogger.debug(
+                "project_generator.download_cache_for_template.no_cache_found",
+                template=template,
+                event_info=("No cache found for template, continuing without it."),
+                target_dir=target_dir,
+            )
+        else:
+            capture_exception_with_context(
+                e,
+                "project_generator.download_cache_for_template.response_error",
+                extra={
+                    "template": template.value,
+                    "status": str(e.status),
+                    "target_dir": target_dir,
+                },
+            )
+    except Exception as exc:
+        capture_exception_with_context(
+            exc,
+            "project_generator.download_cache_for_template.unexpected_error",
+            extra={"template": template.value, "target_dir": target_dir},
+        )
+    finally:
+        # Clean up the temporary file
+        try:
+            Path(temporary_cache_file.name).unlink(missing_ok=True)
+        except Exception as exc:
+            structlogger.debug(
+                "project_generator.download_cache_for_template.cleanup_error",
+                error=str(exc),
+                template=template,
+                event_info=("Failed to cleanup cache for template, ignoring."),
+            )
+
+
+async def background_download_template_caches(
+    app: Sanic, loop: asyncio.AbstractEventLoop
+) -> None:
+    """Kick off background downloads of template caches if enabled."""
+    try:
+        structlogger.info(
+            "builder.main.background_cache_download.start",
+            event_info=(
+                "Starting background download of template caches for this " "version"
+            ),
+        )
+
+        # Ensure cache root exists
+        _cache_root_dir().mkdir(parents=True, exist_ok=True)
+
+        async def _download(template: ProjectTemplateName) -> None:
+            try:
+                target_dir = _template_cache_dir(template)
+                if target_dir.exists() and any(target_dir.iterdir()):
+                    structlogger.debug(
+                        "builder.main.background_cache_download.skipped",
+                        template=template,
+                        event_info=(
+                            "Skipping download of template cache because it "
+                            "already exists."
+                        ),
+                        target_dir=target_dir,
+                    )
+                    return
+
+                target_dir.mkdir(parents=True, exist_ok=True)
+                await download_cache_for_template(template, target_dir.as_posix())
+            except Exception as exc:
+                structlogger.debug(
+                    "builder.main.background_cache_download.error",
+                    template=template,
+                    error=str(exc),
+                )
+
+        # schedule downloads concurrently without blocking startup
+        for template in ProjectTemplateName:
+            loop.create_task(_download(template))
+    except Exception as exc:
+        structlogger.debug(
+            "builder.main.background_cache_download.unexpected_error",
+            error=str(exc),
+        )
+
+
+def copy_cache_for_template_if_available(
+    template: ProjectTemplateName, project_folder: Path
+) -> None:
+    """Copy a previously downloaded cache for `template` into `project_folder`.
+
+    If the cache does not exist, this function is a no-op.
+    """
+    try:
+        cache_dir = _template_cache_dir(template)
+        if cache_dir.exists() and any(cache_dir.iterdir()):
+            _copytree(cache_dir, project_folder)
+            structlogger.info(
+                "project_generator.copy_cache_for_template.success",
+                template=template,
+                event_info=(
+                    "Copied cached template files from cache to project folder."
+                ),
+            )
+        else:
+            structlogger.debug(
+                "project_generator.copy_cache_for_template.missing",
+                template=template,
+                event_info=("No local cache found for template; skipping copy."),
+            )
+    except Exception as exc:
+        structlogger.warning(
+            "project_generator.copy_cache_for_template.error",
+            error=str(exc),
+            template=template,
+        )

rasa/cli/project_templates/basic/config.yml

@@ -11,7 +11,7 @@ assistant_id: placeholder_default
 pipeline:
 - name: SearchReadyLLMCommandGenerator
   llm:
-    model_group: rasa-llm-proxy-openai-llm
+    model_group: openai-gpt-4o
   flow_retrieval:
     active: false
 
@@ -23,8 +23,7 @@ policies:
     type: "faiss"
     source: "./docs"
   llm:
-    model_group: rasa-llm-proxy-openai-llm
+    model_group: openai-gpt-4o
   embeddings:
-    model_group: rasa-llm-proxy-openai-embeddings
-
+    model_group: openai-embeddings
   check_relevancy: true

rasa/cli/project_templates/basic/endpoints.yml

@@ -49,7 +49,7 @@ action_endpoint:
 nlg:
   type: rephrase
   llm:
-    model_group: rasa-llm-proxy-openai-llm
+    model_group: openai-gpt-4o
   prompt_template: prompts/rephraser_demo_personality_prompt.jinja2
   rephrase_all: False
   summarize_history: False
@@ -61,17 +61,7 @@ model_groups:
         model: gpt-4o-2024-11-20
         timeout: 7
         max_tokens: 256
-  - id: rasa-llm-proxy-openai-llm
-    models:
-      - provider: openai
-        api_base: https://hello-llm-proxy.rasa-e2e.workers.dev
-        api_key: ${RASA_PRO_LICENSE}
-        model: gpt-4o-2024-11-20
-        timeout: 7
-        max_tokens: 256
-  - id: rasa-llm-proxy-openai-embeddings
+  - id: openai-embeddings
     models:
       - provider: openai
         model: text-embedding-3-large
-        api_base: https://hello-llm-proxy.rasa-e2e.workers.dev/embeddings
-        api_key: ${RASA_PRO_LICENSE}

rasa/cli/project_templates/default/config.yml

@@ -9,13 +9,13 @@ language: en
 pipeline:
 - name: CompactLLMCommandGenerator
   llm:
-    model_group: rasa-llm-proxy-openai-llm
+    model_group: openai-gpt-4o
 
 # Configuration for Rasa Core.
 policies:
 - name: FlowPolicy
 - name: IntentlessPolicy
   llm:
-    model_group: rasa-llm-proxy-openai-llm
+    model_group: openai-gpt-4o
   embeddings:
-    model_group: rasa-llm-proxy-openai-embeddings
+    model_group: openai-embeddings

rasa/cli/project_templates/default/endpoints.yml

@@ -56,18 +56,7 @@ model_groups:
         model: gpt-4o-2024-11-20
         request_timeout: 7
         max_tokens: 256
-  - id: rasa-llm-proxy-openai-llm
-    models:
-      - provider: openai
-        api_base: https://hello-llm-proxy.rasa-e2e.workers.dev
-        api_key: ${RASA_PRO_LICENSE}
-        model: gpt-4o-2024-11-20
-        timeout: 7
-        max_tokens: 256
-  - id: rasa-llm-proxy-openai-embeddings
+  - id: openai-embeddings
     models:
       - provider: openai
         model: text-embedding-3-large
-        api_base: https://hello-llm-proxy.rasa-e2e.workers.dev/embeddings
-        api_key: ${RASA_PRO_LICENSE}
-

rasa/cli/project_templates/finance/config.yml

@@ -5,7 +5,7 @@ assistant_id: placeholder_default
 pipeline:
 - name: SearchReadyLLMCommandGenerator
   llm:
-    model_group: rasa-llm-proxy-openai-llm
+    model_group: openai-gpt-4o
   flow_retrieval:
     active: false
 
@@ -17,7 +17,7 @@ policies:
     type: "faiss"
     source: "./docs"
   llm:
-    model_group: rasa-llm-proxy-openai-llm
+    model_group: openai-gpt-4o
   embeddings:
-    model_group: rasa-llm-proxy-openai-embeddings
+    model_group: openai-embeddings
   check_relevancy: true

rasa/cli/project_templates/finance/endpoints.yml

@@ -49,7 +49,7 @@ action_endpoint:
 nlg:
   type: rephrase
   llm:
-    model_group: rasa-llm-proxy-openai-llm
+    model_group: openai-gpt-4o
   prompt: prompts/rephraser_demo_personality_prompt.jinja2
   rephrase_all: False
 
@@ -60,17 +60,7 @@ model_groups:
         model: gpt-4o-2024-11-20
         request_timeout: 7
         max_tokens: 256
-  - id: rasa-llm-proxy-openai-llm
-    models:
-      - provider: openai
-        api_base: https://hello-llm-proxy.rasa-e2e.workers.dev
-        api_key: ${RASA_PRO_LICENSE}
-        model: gpt-4o-2024-11-20
-        timeout: 7
-        max_tokens: 256
-  - id: rasa-llm-proxy-openai-embeddings
+  - id: openai-embeddings
     models:
       - provider: openai
         model: text-embedding-3-large
-        api_base: https://hello-llm-proxy.rasa-e2e.workers.dev/embeddings
-        api_key: ${RASA_PRO_LICENSE}