QdrantVectorSearchTool - qdrant_config validation error in Crew Studio

CrewAI Community Support
,
1

Hi,

I’m experiencing a validation error when using QdrantVectorSearchTool in Crew Studio (no-code interface).

Issue:
The no-code interface generates QdrantVectorSearchTool with separate parameters (qdrant_url, qdrant_api_key), but the installed crewai_tools version requires a qdrant_config object.

Error:

pydantic_core._pydantic_core.ValidationError: 1 validation error for QdrantVectorSearchTool
qdrant_config
  Field required [type=missing, input_value={'limit': 5, 'qdrant_url'...ame': 'my_collection'}, input_type=dict]

Generated code (from crew.py):

QdrantVectorSearchTool(limit=5, qdrant_url="https://my-qdrant-server.com", qdrant_api_key="...", collection_name="my_collection")

Expected code:

QdrantVectorSearchTool(
    qdrant_config=QdrantConfig(url="...", api_key="..."),
    collection_name="my_collection",
    limit=5
)

Steps to reproduce:

  1. Create an agent in CrewAI Studio Cloud
  2. Add QdrantVectorSearchTool to the agent
  3. Fill in Qdrant URL, API Key, and Collection Name fields
  4. Run the workflow

Environment variables configured:

  • QDRANT_URL
  • QDRANT_API_KEY

The no-code interface does not allow me to modify the generated Python code directly.

Could you please advise on a workaround or fix this interface-to-library incompatibility?

Thank you.