Documentation
¶
Overview ¶
Package knowledge defines the abstraction for knowledge retrieval providers.
It decouples Cortex from specific RAG engines (e.g., Weave) the same way the llm.Client interface decouples Cortex from specific LLM providers.
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type CollectionInfo ¶
type CollectionInfo struct {
ID string `json:"id"`
Name string `json:"name"`
DocumentCount int64 `json:"document_count"`
ChunkCount int64 `json:"chunk_count"`
EmbeddingModel string `json:"embedding_model,omitempty"`
}
CollectionInfo describes an available knowledge collection.
type Provider ¶
type Provider interface {
// Retrieve performs a semantic search and returns scored chunks.
Retrieve(ctx context.Context, query string, params *RetrieveParams) ([]ScoredChunk, error)
// ListCollections returns available knowledge collections.
ListCollections(ctx context.Context) ([]CollectionInfo, error)
}
Provider is the interface for knowledge retrieval backends. Implementations wrap specific engines (e.g., Weave) to provide knowledge to Cortex agents.
type RetrieveParams ¶
type RetrieveParams struct {
Collection string // Collection name or ID to search in (empty = all).
TopK int // Max results (default: 5).
MinScore float64 // Minimum relevance score (default: 0.0).
}
RetrieveParams configures a knowledge retrieval query.
type ScoredChunk ¶
type ScoredChunk struct {
Content string `json:"content"`
Score float64 `json:"score"`
Source string `json:"source,omitempty"`
DocumentID string `json:"document_id,omitempty"`
CollectionID string `json:"collection_id,omitempty"`
Metadata map[string]string `json:"metadata,omitempty"`
}
ScoredChunk represents a retrieved knowledge chunk with relevance score.
Source Files
Directories
Click to show internal directories.
Click to hide internal directories.