Mythosia.AI.Rag.Abstractions

Mythosia.AI.Rag.Abstractions

Package Summary

Core interfaces and models for the Mythosia.AI RAG ecosystem.
This package defines the contracts that all RAG components implement — you only need this directly if you're building a custom implementation.

Interfaces

Interface	Description
IRagPipeline	Main pipeline contract: ProcessAsync(query) → RagProcessedQuery
IEmbeddingProvider	Text → vector embedding (GetEmbeddingAsync, GetEmbeddingsAsync)
IVectorStore	Vector storage & search (UpsertAsync, SearchAsync, DeleteAsync)
IRagDiagnosticsStore	Optional diagnostics contract (ListAllRecordsAsync, ScoredListAsync)
ITextSplitter	Document → chunks (Split(RagDocument))
IContextBuilder	Search results → LLM prompt (BuildContext(query, results))
IQueryRewriter	Rewrites queries into retrieval-ready form using conversation history, and decides whether document search is needed (search gate)
IRetrievalStrategy	Abstracts retrieval logic — pure vector or hybrid (BM25 + vector + RRF)
IReranker	Re-ranks search results post-retrieval for improved relevance

Models

Model	Description
RagChunk	A chunk of text with ID, content, document ID, index, and metadata
RagDocument	A loaded document with Id, Content, Source, and Metadata for the RAG pipeline
RagProcessedQuery	Pipeline output: original query, rewritten semantic query, retrieval keywords, RequestMessageContent, references, RetrievalCandidates, SearchSkipped, RewriteResult, HasReferences flag, and Diagnostics
QueryRewriteResult	Result of rewriting a query into retrieval-ready form, including search gate decision (NeedsSearch) and optional retrieval keywords. Factory methods Pass() and Search()
ConversationTurn	Lightweight DTO representing a single conversation turn (role + content) for IQueryRewriter context
RagQueryDiagnostics	Applied retrieval metadata (FinalTopK, RetrievalTopK, AppliedFinalMinScore, AppliedRetrievalMinScore, ElapsedMs, RewriteElapsedMs)
RagPipelineOptions	Configuration: DefaultQuery, PromptTemplate, EmbeddingBatchSize. Includes Clone() for snapshot-style runtime updates
RagQueryOptions	Per-request overrides: FinalFilter, RetrievalDerivation, StoreFilter, FinalSelection, ProgressAsync. Includes Clone() for safe per-query overrides on top of RagPipelineOptions.DefaultQuery
RagFinalSelectionOptions	Final selection policy after re-ranking (Mode, RetrievalWeight). Includes Clone()
RagFinalSelectionMode	Enum: RerankerOnly (default) or WeightedBlend
RagFilter	Final selection policy (TopK, MinScore). Includes Clone()
RagRetrievalDerivation	Controls how retrieval candidates are derived (TopKMultiplier, MinScoreDivider). Includes Clone()
RagRetrievalFilter	Immutable computed retrieval filter (TopK, MinScore)
RagProgressStage	Enum for pipeline stage progress reporting (QueryRewrite, Embedding, Filtering, Retrieval, Reranking, ContextBuild)
VectorRecord	Stored vector with ID, content, embedding, and metadata
VectorSearchResult	Search result with record and similarity score
VectorFilter	Filter by metadata conditions or minimum score

Per-Query Overrides via Clone()

RagPipelineOptions, RagQueryOptions, RagFilter, RagRetrievalDerivation, and RagFinalSelectionOptions all expose a Clone() method. Use Clone() to override a single field on top of RagPipelineOptions.DefaultQuery without losing the other configured fields (notably StoreFilter and ProgressAsync):

var options = pipeline.Options.DefaultQuery.Clone();
options.FinalFilter.TopK = 10;
await pipeline.QueryAsync(query, options);

Constructing new RagQueryOptions { FinalFilter = ... } from scratch silently drops every other field, including any tenant/permission scope on StoreFilter and any progress callback on ProgressAsync. Clone() makes the "inherit defaults, override one field" pattern safe.

Custom Implementation Example

public class MyEmbeddingProvider : IEmbeddingProvider
{
    public int Dimensions => 768;

    public Task<float[]> GetEmbeddingAsync(string text, CancellationToken ct = default)
    {
        // Your embedding logic here
    }

    public Task<IReadOnlyList<float[]>> GetEmbeddingsAsync(
        IEnumerable<string> texts, CancellationToken ct = default)
    {
        // Batch embedding logic here
    }
}

Then register via the builder:

.WithRag(rag => rag.UseEmbedding(new MyEmbeddingProvider()))