microsoft/autogen · error · ValueError

vector_fields must be provided for vector search

Error message

vector_fields must be provided for vector search

What it means

The config model's second interdependent rule: query_type='vector' requires a non-empty vector_fields list, because the client must know which vector field of the index receives the query embedding. It backs up the factory-level checks (errors 926/927) and also applies to directly constructed or deserialized configs.

Source

Thrown at python/packages/autogen-ext/src/autogen_ext/tools/azure/_config.py:176

            return "full"

        return v

    @field_validator("top")
    def validate_top(cls, v: Optional[int]) -> Optional[int]:
        """Ensure top is a positive integer if provided."""
        if v is not None and v <= 0:
            raise ValueError("top must be a positive integer")
        return v

    @model_validator(mode="after")
    def validate_interdependent_fields(self) -> "AzureAISearchConfig":
        """Validate interdependent fields after all fields have been parsed."""
        if self.query_type == "semantic" and not self.semantic_config_name:
            raise ValueError("semantic_config_name must be provided when query_type is 'semantic'")

        if self.query_type == "vector" and not self.vector_fields:
            raise ValueError("vector_fields must be provided for vector search")

        if (
            self.embedding_provider
            and self.embedding_provider.lower() == "azure_openai"
            and self.embedding_model
            and not self.openai_endpoint
        ):
            raise ValueError("openai_endpoint must be provided for azure_openai embedding provider")

        return self

View on GitHub (pinned to 027ecf0a37)

Solutions

1. Set vector_fields=['<vector-field-name>'] matching the index's vector field (Collection(Edm.Single)).
2. Or change query_type to 'simple'/'full' if you only need text search.
3. Validate saved configs in CI by constructing AzureAISearchConfig from the file contents.

Example fix

# before
AzureAISearchConfig(name='s', endpoint=ep, index_name='idx', credential=cred, query_type='vector')
# after
AzureAISearchConfig(name='s', endpoint=ep, index_name='idx', credential=cred, query_type='vector', vector_fields=['content_vector'])

def vector_fields_rule_ok(cfg_dict: dict) -> bool:
    if cfg_dict.get('query_type') == 'vector':
        vf = cfg_dict.get('vector_fields')
        return isinstance(vf, (list, tuple)) and len(vf) > 0
    return True

def vector_fields_present(query_type: str, vector_fields) -> bool:
    return query_type != 'vector' or (isinstance(vector_fields, (list, tuple)) and len(vector_fields) > 0)

When it happens

Trigger: AzureAISearchConfig(query_type='vector') (or deserialized config with that query_type) where vector_fields is None, missing, or []; note also that hybrid requires vector_fields via the factory path even though this specific validator keys on 'vector'.

Common situations: Hand-editing serialized component configs to enable vector search without adding vector_fields; configs from an older schema version; the index's vector field renamed so the list was emptied.

Related errors

• Invalid configuration: {str(e)}
• vector_fields must contain at least one field name for vecto
• vector_fields must contain at least one field name for hybri
• endpoint must be a valid URL starting with http:// or https:
• top must be a positive integer

AI-assisted analysis of microsoft/autogen@027ecf0a37 (2026-08-15). Data as JSON: /api/errors/306a1bf750eb55e1. Report an issue: GitHub.