deepset-mcp 0.0.2rc1__tar.gz → 0.0.3__tar.gz

{deepset_mcp-0.0.2rc1 → deepset_mcp-0.0.3}/.github/workflows/pypi_release.yml

@@ -12,14 +12,12 @@ jobs:
     - uses: actions/checkout@v4
       with:
         persist-credentials: false
-    - name: Set up Python
-      uses: actions/setup-python@v5
-      with:
-        python-version: "3.x"
     - name: Install uv
-      run: >-
-        curl -LsSf https://astral.sh/uv/install.sh | sh &&
-        echo "$HOME/.cargo/bin" >> $GITHUB_PATH
+      uses: astral-sh/setup-uv@v5
+    - name: Set up Python
+      run: uv python install
+    - name: Install the project
+      run: uv sync --locked --all-extras --dev
     - name: Build a binary wheel and a source tarball
       run: uv build
     - name: Store the distribution packages
@@ -47,24 +45,3 @@ jobs:
         path: dist/
     - name: Publish distribution to PyPI
       uses: pypa/gh-action-pypi-publish@release/v1
-  publish-to-testpypi:
-    name: Publish  deepset-mcp distribution to TestPyPI
-    if: startsWith(github.ref, 'refs/tags/')  # only publish to TestPyPI on tag pushes
-    needs:
-    - build
-    runs-on: ubuntu-latest
-    environment:
-      name: testpypi
-      url: https://test.pypi.org/p/deepset-mcp
-    permissions:
-      id-token: write  # IMPORTANT: mandatory for trusted publishing
-    steps:
-    - name: Download all the dists
-      uses: actions/download-artifact@v4
-      with:
-        name: python-package-distributions
-        path: dist/
-    - name: Publish distribution to TestPyPI
-      uses: pypa/gh-action-pypi-publish@release/v1
-      with:
-        repository-url: https://test.pypi.org/legacy/

{deepset_mcp-0.0.2rc1 → deepset_mcp-0.0.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: deepset-mcp
-Version: 0.0.2rc1
+Version: 0.0.3
 Summary: Collection of MCP tools and Agents to work with the deepset AI platform. Create, debug or learn about pipelines on the platform. Useable from the CLI, Cursor, Claude Code, or other MCP clients.
 Project-URL: Homepage, https://deepset.ai
 Author-email: Mathis Lucka <mathis.lucka@deepset.ai>, Tanay Soni <tanay.soni@deepset.ai>
@@ -18,10 +18,11 @@ Requires-Python: >=3.11
 Requires-Dist: fastapi
 Requires-Dist: glom
 Requires-Dist: httpx
-Requires-Dist: mcp
+Requires-Dist: mcp>=1.10.1
 Requires-Dist: model2vec
 Requires-Dist: numpy
 Requires-Dist: pydantic>=2.0.0
+Requires-Dist: pyjwt[crypto]
 Requires-Dist: pyyaml
 Requires-Dist: rich
 Provides-Extra: analysis
@@ -117,18 +118,14 @@ Agents can use these tools to:
 Running the server with uv gives you faster startup time and consumes slightly less resources on your system.
 
 1. [Install uv](https://docs.astral.sh/uv/guides/install-python/) if you don't have it yet
-2. Clone the [deepset-mcp-server repository](https://github.com/deepset-ai/deepset-mcp-server)
-3. Put the following into your `claude_desktop_config.json`
+2. Put the following into your `claude_desktop_config.json`
 
 ```python
 {
   "mcpServers": {
     "deepset": {
-      "command": "/opt/homebrew/bin/uv", # path to your uv installation
+      "command": "uvx",
       "args": [
-        "--directory",
-        "/path/to/your/clone/of/deepset-mcp-server", # path to your clone of the deepset-mcp-server repo 
-        "run",
         "deepset-mcp"
       ],
       "env": {
@@ -141,7 +138,10 @@ Running the server with uv gives you faster startup time and consumes slightly l
 }
 ```
 
-4. Quit and start the Claude Desktop App
+This will load the [deepset-mcp package from PyPi](https://pypi.org/project/deepset-mcp/) and install it into a temporary virtual environment.
+
+3. Quit and start the Claude Desktop App
+
 
 
 ### Other MCP Clients
@@ -156,7 +156,7 @@ Here is where you need to configure `deepset-mcp` for:
 
 Generally speaking, depending on your installation, you need to configure an MCP client with one of the following commands:
 
-`uv --directory path/to/deepset-mcp run deepset-mcp --workspace your_workspace --api-key your_api_key`
+`uvx deepset-mcp --workspace your_workspace --api-key your_api_key`
 
 If you installed the deepset-mcp package globally and added it to your `PATH`, you can just run:
 
@@ -192,11 +192,8 @@ For example:
 {
   "mcpServers": {
     "deepset": {
-      "command": "/opt/homebrew/bin/uv",
+      "command": "uvx",
       "args": [
-        "--directory",
-        "/path/to/your/clone/of/deepset-mcp-server",
-        "run",
         "deepset-mcp",
         "--workspace-mode",
         "explicit"
@@ -272,7 +269,7 @@ If your pipeline is not deployed yet, the LLM can autonomously validate it and f
 ## CLI
 You can use the MCP server as a Haystack Agent through a command-line interface.
 
-Install with `uv pip install deepset-mcp[cli]`.
+Install with `uvx tool install "deepset-mcp[cli]"`.
 
 Start the interactive CLI with:
 

{deepset_mcp-0.0.2rc1 → deepset_mcp-0.0.3}/README.md

@@ -78,18 +78,14 @@ Agents can use these tools to:
 Running the server with uv gives you faster startup time and consumes slightly less resources on your system.
 
 1. [Install uv](https://docs.astral.sh/uv/guides/install-python/) if you don't have it yet
-2. Clone the [deepset-mcp-server repository](https://github.com/deepset-ai/deepset-mcp-server)
-3. Put the following into your `claude_desktop_config.json`
+2. Put the following into your `claude_desktop_config.json`
 
 ```python
 {
   "mcpServers": {
     "deepset": {
-      "command": "/opt/homebrew/bin/uv", # path to your uv installation
+      "command": "uvx",
       "args": [
-        "--directory",
-        "/path/to/your/clone/of/deepset-mcp-server", # path to your clone of the deepset-mcp-server repo 
-        "run",
         "deepset-mcp"
       ],
       "env": {
@@ -102,7 +98,10 @@ Running the server with uv gives you faster startup time and consumes slightly l
 }
 ```
 
-4. Quit and start the Claude Desktop App
+This will load the [deepset-mcp package from PyPi](https://pypi.org/project/deepset-mcp/) and install it into a temporary virtual environment.
+
+3. Quit and start the Claude Desktop App
+
 
 
 ### Other MCP Clients
@@ -117,7 +116,7 @@ Here is where you need to configure `deepset-mcp` for:
 
 Generally speaking, depending on your installation, you need to configure an MCP client with one of the following commands:
 
-`uv --directory path/to/deepset-mcp run deepset-mcp --workspace your_workspace --api-key your_api_key`
+`uvx deepset-mcp --workspace your_workspace --api-key your_api_key`
 
 If you installed the deepset-mcp package globally and added it to your `PATH`, you can just run:
 
@@ -153,11 +152,8 @@ For example:
 {
   "mcpServers": {
     "deepset": {
-      "command": "/opt/homebrew/bin/uv",
+      "command": "uvx",
       "args": [
-        "--directory",
-        "/path/to/your/clone/of/deepset-mcp-server",
-        "run",
         "deepset-mcp",
         "--workspace-mode",
         "explicit"
@@ -233,7 +229,7 @@ If your pipeline is not deployed yet, the LLM can autonomously validate it and f
 ## CLI
 You can use the MCP server as a Haystack Agent through a command-line interface.
 
-Install with `uv pip install deepset-mcp[cli]`.
+Install with `uvx tool install "deepset-mcp[cli]"`.
 
 Start the interactive CLI with:
 

{deepset_mcp-0.0.2rc1 → deepset_mcp-0.0.3}/pyproject.toml

@@ -37,7 +37,7 @@ requires-python = ">=3.11"
 
 dependencies = [
     "fastapi",
-    "mcp",
+    "mcp>=1.10.1",
     "httpx",
     "pydantic>=2.0.0",
     "pyyaml",
@@ -45,6 +45,7 @@ dependencies = [
     "model2vec",
     "glom",
     "rich",
+    "pyjwt[crypto]",
 ]
 
 [project.urls]

{deepset_mcp-0.0.2rc1 → deepset_mcp-0.0.3}/src/deepset_mcp/api/pipeline_template/models.py

@@ -1,7 +1,8 @@
 from enum import StrEnum
+from typing import Any
 from uuid import UUID
 
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, model_validator
 
 
 class PipelineType(StrEnum):
@@ -28,10 +29,36 @@ class PipelineTemplate(BaseModel):
     display_name: str = Field(alias="name")
     pipeline_template_id: UUID = Field(alias="pipeline_template_id")
     potential_applications: list[str] = Field(alias="potential_applications")
-    yaml_config: str | None = Field(None, alias="query_yaml")
+    yaml_config: str | None = None
     tags: list[PipelineTemplateTag]
     pipeline_type: PipelineType
 
+    @model_validator(mode="before")
+    @classmethod
+    def populate_yaml_config(cls, values: Any) -> Any:
+        """Populate yaml_config from query_yaml or indexing_yaml based on pipeline_type."""
+        if not isinstance(values, dict):
+            return values
+
+        # Skip if yaml_config is already set
+        if values.get("yaml_config") is not None:
+            return values
+
+        # Get pipeline_type from the model data
+        pipeline_type = values.get("pipeline_type")
+
+        if pipeline_type == PipelineType.INDEXING or pipeline_type == "indexing":
+            yaml_config = values.get("indexing_yaml")
+        elif pipeline_type == PipelineType.QUERY or pipeline_type == "query":
+            yaml_config = values.get("query_yaml")
+        else:
+            yaml_config = None
+
+        if yaml_config is not None:
+            values["yaml_config"] = yaml_config
+
+        return values
+
 
 class PipelineTemplateList(BaseModel):
     """Response model for listing pipeline templates."""

deepset_mcp-0.0.3/src/deepset_mcp/config.py

@@ -0,0 +1,25 @@
+"""This module contains static configuration for the deepset MCP server."""
+
+# We need this mapping to which environment variables integrations are mapped to
+# The mapping is maintained in the pipeline operator:
+# https://github.com/deepset-ai/dc-pipeline-operator/blob/main/dc_operators/config.py#L279
+TOKEN_DOMAIN_MAPPING = {
+    "huggingface.co": ["HF_API_TOKEN", "HF_TOKEN"],
+    "api.openai.com": ["OPENAI_API_KEY"],
+    "bedrock.amazonaws.com": ["BEDROCK"],
+    "api.cohere.ai": ["COHERE_API_KEY"],
+    "openai.azure.com": ["AZURE_OPENAI_API_KEY"],
+    "cognitive-services.azure.com": ["AZURE_AI_API_KEY"],
+    "unstructured.io": ["UNSTRUCTURED_API_KEY"],
+    "api.deepl.com": ["DEEPL_API_KEY"],
+    "generativelanguage.googleapis.com": ["GOOGLE_API_KEY"],
+    "api.nvidia.com": ["NVIDIA_API_KEY"],
+    "api.voyageai.com": ["VOYAGE_API_KEY"],
+    "searchapi.io": ["SEARCHAPI_API_KEY"],
+    "snowflakecomputing.com": ["SNOWFLAKE_API_KEY"],
+    "wandb.ai": ["WANDB_API_KEY"],
+    "mongodb.com": ["MONGO_CONNECTION_STRING"],
+    "together.ai": ["TOGETHERAI_API_KEY"],
+}
+
+DEEPSET_DOCS_DEFAULT_SHARE_URL = "https://cloud.deepset.ai/shared_prototypes?share_token=prototype_eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJleHAiOjE3ODM0MjE0OTguNTk5LCJhdWQiOiJleHRlcm5hbCB1c2VyIiwiaXNzIjoiZEMiLCJ3b3Jrc3BhY2VfaWQiOiI4YzI0ZjExMi1iMjljLTQ5MWMtOTkzOS1hZTkxMDRhNTQyMWMiLCJ3b3Jrc3BhY2VfbmFtZSI6ImRjLWRvY3MtY29udGVudCIsIm9yZ2FuaXphdGlvbl9pZCI6ImNhOWYxNGQ0LWMyYzktNDYwZC04ZDI2LWY4Y2IwYWNhMDI0ZiIsInNoYXJlX2lkIjoiY2Y3MTA3ODAtOThmNi00MzlmLThiNzYtMmMwNDkyODNiMDZhIiwibG9naW5fcmVxdWlyZWQiOmZhbHNlfQ.5j6DCNRQ1_KB8lhIJqHyw2hBIleEW1_Y_UBuH6MTYY0"

deepset_mcp-0.0.3/src/deepset_mcp/main.py

@@ -0,0 +1,184 @@
+import argparse
+import asyncio
+import logging
+import os
+from pathlib import Path
+from urllib.parse import parse_qs, urlparse
+
+import jwt
+from mcp.server.fastmcp import FastMCP
+
+from deepset_mcp.api.client import AsyncDeepsetClient
+from deepset_mcp.config import DEEPSET_DOCS_DEFAULT_SHARE_URL
+from deepset_mcp.tool_factory import WorkspaceMode, register_tools
+
+# Initialize MCP Server
+mcp = FastMCP("Deepset Cloud MCP", settings={"log_level": "ERROR"})
+
+logging.getLogger("uvicorn").setLevel(logging.WARNING)
+logging.getLogger("uvicorn.access").setLevel(logging.WARNING)
+logging.getLogger("fastapi").setLevel(logging.WARNING)
+logging.getLogger("httpx").setLevel(logging.WARNING)
+logging.getLogger("httpcore").setLevel(logging.WARNING)
+logging.getLogger("mcp").setLevel(logging.WARNING)
+
+
+@mcp.prompt()
+async def deepset_copilot() -> str:
+    """System prompt for the deepset copilot."""
+    prompt_path = Path(__file__).parent / "prompts/deepset_copilot_prompt.md"
+
+    return prompt_path.read_text()
+
+
+@mcp.prompt()
+async def deepset_recommended_prompt() -> str:
+    """Recommended system prompt for the deepset copilot."""
+    prompt_path = Path(__file__).parent / "prompts/deepset_debugging_agent.md"
+
+    return prompt_path.read_text()
+
+
+async def fetch_shared_prototype_details(share_url: str) -> tuple[str, str, str]:
+    """Gets the pipeline name, workspace name and an API token for a shared prototype url.
+
+    :param share_url: The URL of a shared prototype on the deepset platform.
+
+    :returns: A tuple containing the pipeline name, workspace name and an API token.
+    """
+    parsed_url = urlparse(share_url)
+    query_params = parse_qs(parsed_url.query)
+    share_token = query_params.get("share_token", [None])[0]
+    if not share_token:
+        raise ValueError("Invalid share URL: missing share_token parameter.")
+
+    jwt_token = share_token.replace("prototype_", "")
+
+    decoded_token = jwt.decode(jwt_token, options={"verify_signature": False})
+    workspace_name = decoded_token.get("workspace_name")
+    if not workspace_name:
+        raise ValueError("Invalid JWT in share_token: missing 'workspace_name'.")
+
+    share_id = decoded_token.get("share_id")
+    if not share_id:
+        raise ValueError("Invalid JWT in share_token: missing 'share_id'.")
+
+    # For shared prototypes, we need to:
+    # 1. Fetch prototype details (pipeline name) using the information encoded in the JWT
+    # 2. Create a shared prototype user
+    async with AsyncDeepsetClient(api_key=share_token) as client:
+        response = await client.request(f"/v1/workspaces/{workspace_name}/shared_prototypes/{share_id}")
+        if not response.success:
+            raise ValueError(f"Failed to fetch shared prototype details: {response.status_code} {response.json}")
+
+        data = response.json or {}
+        pipeline_names: list[str] = data.get("pipeline_names", [])
+        if not pipeline_names:
+            raise ValueError("No pipeline names found in shared prototype response.")
+
+        user_info = await client.request("/v1/workspaces/dc-docs-content/shared_prototype_users", method="POST")
+
+        if not user_info.success:
+            raise ValueError("Failed to fetch user information from shared prototype response.")
+
+        user_data = user_info.json or {}
+
+        try:
+            api_key = user_data["user_token"]
+        except KeyError:
+            raise ValueError("No user token in shared prototype response.") from None
+
+    return workspace_name, pipeline_names[0], api_key
+
+
+def main() -> None:
+    """Entrypoint for the deepset MCP server."""
+    parser = argparse.ArgumentParser(description="Run the Deepset MCP server.")
+    parser.add_argument(
+        "--workspace",
+        "-w",
+        help="Deepset workspace (env DEEPSET_WORKSPACE)",
+    )
+    parser.add_argument(
+        "--api-key",
+        "-k",
+        help="Deepset API key (env DEEPSET_API_KEY)",
+    )
+    parser.add_argument(
+        "--docs-share-url",
+        default=DEEPSET_DOCS_DEFAULT_SHARE_URL,
+        help="Deepset docs search share URL (env DEEPSET_DOCS_SHARE_URL)",
+    )
+    parser.add_argument(
+        "--workspace-mode",
+        choices=[WorkspaceMode.STATIC, WorkspaceMode.DYNAMIC],
+        default=WorkspaceMode.STATIC,
+        help=(
+            "Whether workspace should be set statically or dynamically provided during a tool call. "
+            f"Default: '{WorkspaceMode.STATIC}'"
+        ),
+    )
+    parser.add_argument(
+        "--tools",
+        nargs="*",
+        help="Space-separated list of tools to register (default: all)",
+    )
+    parser.add_argument(
+        "--list-tools",
+        action="store_true",
+        help="List all available tools and exit",
+    )
+    args = parser.parse_args()
+
+    # Handle --list-tools flag early
+    if args.list_tools:
+        from deepset_mcp.tool_factory import TOOL_REGISTRY
+
+        print("Available tools:")
+        for tool_name in sorted(TOOL_REGISTRY.keys()):
+            print(f"  {tool_name}")
+        return
+
+    # prefer flags, fallback to env
+    workspace = args.workspace or os.getenv("DEEPSET_WORKSPACE")
+    api_key = args.api_key or os.getenv("DEEPSET_API_KEY")
+    docs_share_url = args.docs_share_url or os.getenv("DEEPSET_DOCS_SHARE_URL")
+
+    if docs_share_url:
+        try:
+            workspace_name, pipeline_name, api_key_docs = asyncio.run(fetch_shared_prototype_details(docs_share_url))
+            os.environ["DEEPSET_DOCS_WORKSPACE"] = workspace_name
+            os.environ["DEEPSET_DOCS_PIPELINE_NAME"] = pipeline_name
+            os.environ["DEEPSET_DOCS_API_KEY"] = api_key_docs
+        except (ValueError, jwt.DecodeError) as e:
+            parser.error(f"Error processing --docs-share-url: {e}")
+
+    # Create server configuration
+    workspace_mode = WorkspaceMode(args.workspace_mode)
+
+    if workspace_mode == WorkspaceMode.STATIC:
+        if not workspace:
+            parser.error("Missing workspace: set --workspace or DEEPSET_WORKSPACE")
+
+    if not api_key:
+        parser.error("Missing API key: set --api-key or DEEPSET_API_KEY")
+
+    # make sure downstream tools see them (for implicit mode)
+    if workspace:
+        os.environ["DEEPSET_WORKSPACE"] = workspace
+    os.environ["DEEPSET_API_KEY"] = api_key
+
+    # Parse tool names if provided
+    tool_names = None
+    if args.tools:
+        tool_names = set(args.tools)
+
+    # Register tools based on configuration
+    register_tools(mcp, workspace_mode, workspace, tool_names)
+
+    # run with SSE transport (HTTP+Server-Sent Events)
+    mcp.run(transport="stdio")
+
+
+if __name__ == "__main__":
+    main()

{deepset_mcp-0.0.2rc1 → deepset_mcp-0.0.3}/src/deepset_mcp/tool_factory.py

@@ -19,7 +19,6 @@ from deepset_mcp.tools.custom_components import (
     list_custom_component_installations as list_custom_component_installations_tool,
 )
 from deepset_mcp.tools.doc_search import (
-    get_docs_config,
     search_docs as search_docs_tool,
 )
 from deepset_mcp.tools.haystack_service import (
@@ -48,9 +47,9 @@ from deepset_mcp.tools.pipeline import (
     validate_pipeline as validate_pipeline_tool,
 )
 from deepset_mcp.tools.pipeline_template import (
-    get_pipeline_template as get_pipeline_template_tool,
-    list_pipeline_templates as list_pipeline_templates_tool,
-    search_pipeline_templates as search_pipeline_templates_tool,
+    get_template as get_pipeline_template_tool,
+    list_templates as list_pipeline_templates_tool,
+    search_templates as search_pipeline_templates_tool,
 )
 from deepset_mcp.tools.secrets import (
     get_secret as get_secret_tool,
@@ -63,6 +62,16 @@ from deepset_mcp.tools.workspace import (
     list_workspaces as list_workspaces_tool,
 )
 
+
+def are_docs_available() -> bool:
+    """Checks if documentation search is available."""
+    return bool(
+        os.environ.get("DEEPSET_DOCS_WORKSPACE", False)
+        and os.environ.get("DEEPSET_DOCS_PIPELINE_NAME", False)
+        and os.environ.get("DEEPSET_DOCS_API_KEY", False)
+    )
+
+
 EXPLORER = RichExplorer(store=STORE)
 
 
@@ -108,16 +117,11 @@ async def search_docs(query: str) -> str:
     :param query: The search query to execute against the documentation.
     :returns: The formatted search results from the documentation.
     """
-    docs_config = get_docs_config()
-    if not docs_config:
-        raise RuntimeError("Documentation search configuration not available")
-
-    docs_workspace, docs_pipeline_name, docs_api_key = docs_config
-    async with AsyncDeepsetClient(api_key=docs_api_key) as client:
+    async with AsyncDeepsetClient(api_key=os.environ["DEEPSET_DOCS_API_KEY"]) as client:
         response = await search_docs_tool(
             client=client,
-            workspace=docs_workspace,
-            pipeline_name=docs_pipeline_name,
+            workspace=os.environ["DEEPSET_DOCS_WORKSPACE"],
+            pipeline_name=os.environ["DEEPSET_DOCS_PIPELINE_NAME"],
             query=query,
         )
     return response
@@ -126,8 +130,8 @@ async def search_docs(query: str) -> str:
 class WorkspaceMode(StrEnum):
     """Configuration for how workspace is provided to tools."""
 
-    IMPLICIT = "implicit"  # workspace from env, no parameter in tool signature
-    EXPLICIT = "explicit"  # workspace as required parameter in tool signature
+    STATIC = "static"  # workspace from env, no parameter in tool signature
+    DYNAMIC = "dynamic"  # workspace as required parameter in tool signature
 
 
 class MemoryType(StrEnum):
@@ -165,11 +169,21 @@ TOOL_REGISTRY: dict[str, tuple[Callable[..., Any], ToolConfig]] = {
     ),
     "create_pipeline": (
         create_pipeline_tool,
-        ToolConfig(needs_client=True, needs_workspace=True, memory_type=MemoryType.BOTH),
+        ToolConfig(
+            needs_client=True,
+            needs_workspace=True,
+            memory_type=MemoryType.BOTH,
+            custom_args={"skip_validation_errors": True},
+        ),
     ),
     "update_pipeline": (
         update_pipeline_tool,
-        ToolConfig(needs_client=True, needs_workspace=True, memory_type=MemoryType.BOTH),
+        ToolConfig(
+            needs_client=True,
+            needs_workspace=True,
+            memory_type=MemoryType.BOTH,
+            custom_args={"skip_validation_errors": True},
+        ),
     ),
     "get_pipeline": (
         get_pipeline_tool,
@@ -216,15 +230,20 @@ TOOL_REGISTRY: dict[str, tuple[Callable[..., Any], ToolConfig]] = {
         deploy_index_tool,
         ToolConfig(needs_client=True, needs_workspace=True, memory_type=MemoryType.EXPLORABLE),
     ),
-    "list_pipeline_templates": (
+    "list_templates": (
         list_pipeline_templates_tool,
-        ToolConfig(needs_client=True, needs_workspace=True, memory_type=MemoryType.EXPLORABLE),
+        ToolConfig(
+            needs_client=True,
+            needs_workspace=True,
+            memory_type=MemoryType.EXPLORABLE,
+            custom_args={"field": "created_at", "order": "DESC", "limit": 100},
+        ),
     ),
-    "get_pipeline_template": (
+    "get_template": (
         get_pipeline_template_tool,
         ToolConfig(needs_client=True, needs_workspace=True, memory_type=MemoryType.EXPLORABLE),
     ),
-    "search_pipeline_templates": (
+    "search_templates": (
         search_pipeline_templates_tool,
         ToolConfig(
             needs_client=True,
@@ -290,8 +309,8 @@ def create_enhanced_tool(
     Args:
         base_func: The base tool function.
         config: Tool configuration specifying dependencies and custom arguments.
-        workspace_mode: How the workspace should be handled (implicit or explicit).
-        workspace: The workspace to use for implicit mode.
+        workspace_mode: How the workspace should be handled.
+        workspace: The workspace to use when using a static workspace.
 
     Returns:
         An enhanced, awaitable tool function with an updated signature and docstring.
@@ -333,7 +352,7 @@ def create_enhanced_tool(
         params_to_remove.update(config.custom_args.keys())
     if config.needs_client:
         params_to_remove.add("client")
-    if config.needs_workspace and workspace_mode == WorkspaceMode.IMPLICIT:
+    if config.needs_workspace and workspace_mode == WorkspaceMode.STATIC:
         params_to_remove.add("workspace")
 
     # Create the new signature from the original function
@@ -350,15 +369,15 @@ def create_enhanced_tool(
     # Create the final wrapper function that handles client/workspace injection
     if config.needs_client:
         if config.needs_workspace:
-            if workspace_mode == WorkspaceMode.IMPLICIT:
+            if workspace_mode == WorkspaceMode.STATIC:
 
-                async def workspace_implicit_wrapper(**kwargs: Any) -> Any:
+                async def workspace_environment_wrapper(**kwargs: Any) -> Any:
                     ws = workspace or get_workspace_from_env()
                     async with AsyncDeepsetClient() as client:
                         return await decorated_func(client=client, workspace=ws, **kwargs)
 
-                wrapper = workspace_implicit_wrapper
-            else:  # EXPLICIT mode
+                wrapper = workspace_environment_wrapper
+            else:  # DYNAMIC mode
 
                 async def workspace_explicit_wrapper(**kwargs: Any) -> Any:
                     async with AsyncDeepsetClient() as client:
@@ -400,7 +419,7 @@ def create_enhanced_tool(
         params_to_remove_from_doc = set()
         if config.needs_client:
             params_to_remove_from_doc.add("client")
-        if config.needs_workspace and workspace_mode == WorkspaceMode.IMPLICIT:
+        if config.needs_workspace and workspace_mode == WorkspaceMode.STATIC:
             params_to_remove_from_doc.add("workspace")
         if config.custom_args:
             params_to_remove_from_doc.update(config.custom_args.keys())
@@ -428,11 +447,11 @@ def register_tools(
     Args:
         mcp: FastMCP server instance
         workspace_mode: How workspace should be handled
-        workspace: Workspace to use for implicit mode (if None, reads from env)
+        workspace: Workspace to use for environment mode (if None, reads from env)
         tool_names: Set of tool names to register (if None, registers all tools)
     """
     # Check if docs search is available
-    docs_available = get_docs_config() is not None
+    docs_available = are_docs_available()
 
     # Validate tool names if provided
     if tool_names is not None:
@@ -446,8 +465,8 @@ def register_tools(
         # Warn if search_docs was requested but config is missing
         if "search_docs" in tool_names and not docs_available:
             logging.warning(
-                "Documentation search tool requested but not available. To enable, set the following environment "
-                "variables: DEEPSET_DOCS_WORKSPACE, DEEPSET_DOCS_PIPELINE_NAME, DEEPSET_DOCS_API_KEY"
+                "Documentation search tool requested but not available. To enable, set the DEEPSET_DOCS_SHARE_URL "
+                "environment variable."
             )
 
         tools_to_register = tool_names.copy()
@@ -457,8 +476,7 @@ def register_tools(
         # Warn if search_docs would be skipped in "all tools" mode
         if not docs_available:
             logging.warning(
-                "Documentation search tool not enabled. To enable, set the following environment "
-                "variables: DEEPSET_DOCS_WORKSPACE, DEEPSET_DOCS_PIPELINE_NAME, DEEPSET_DOCS_API_KEY"
+                "Documentation search tool not enabled. To enable, set the DEEPSET_DOCS_SHARE_URL environment variable."
             )
 
     # Remove search_docs if config is not available
@@ -470,4 +488,4 @@ def register_tools(
         # Create enhanced tool
         enhanced_tool = create_enhanced_tool(base_func, config, workspace_mode, workspace)
 
-        mcp.add_tool(enhanced_tool, name=tool_name)
+        mcp.add_tool(enhanced_tool, name=tool_name, structured_output=False)