[dotnet-ai] RAG pipeline - #245
Conversation
Adds the RAG pipeline skill covering the 6-stage retrieval-augmented generation pattern: ingest, embed, store, retrieve, assemble, generate. Composes meai-chat, meai-embeddings, vector-data-search, and data-ingestion. Fixes dotnet#233 Part of dotnet#225 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
.NET 10 is the current LTS (released Nov 2025). .NET 8 reaches end-of-support Nov 2026. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
|
@luisquintanilla - we want skills to be atomic to the extent possible ( see contributing.md ). Tagging @artl93 @joperezr and @jeffhandley to review and decide the course of action. |
|
@ManishJayaswal - can you define what makes this not atomic? It's basically a RAG project template with validation. |
| dotnet add package Microsoft.Extensions.VectorData.Abstractions | ||
|
|
||
| # Dev / prototyping | ||
| dotnet add package Microsoft.Extensions.VectorData.InMemory |
There was a problem hiding this comment.
Is this something that should be parameterized?
| var systemPrompt = $""" | ||
| Answer the user's question based ONLY on the provided context. | ||
| If the context doesn't contain the answer, say "I don't have information about that." | ||
| Always cite your sources. |
There was a problem hiding this comment.
Consider whether this is strong enough as a system prompt - should it be more guarded against prompt injection?
|
This pull request has been automatically closed because it has been open for more than 30 days with no recent activity. If you believe this work is still relevant, please feel free to reopen or create a new pull request. Thank you for your contribution! Warning Firewall blocked 1 domainThe following domain was blocked by the firewall during workflow execution:
network:
allowed:
- defaults
- "awmgmcpg"See Network Configuration for more information.
|
Fixes #233
Part of #225
Dependencies
Summary
Adds the
rag-pipelineskill covering end-to-end retrieval-augmented generation. This is a composition skill that orchestrates the foundation skills (chat, embeddings, vector search, ingestion) into a 6-stage pattern: Ingest Embed Store Retrieve Assemble Generate.Changes
plugins/dotnet-ai/skills/rag-pipeline/SKILL.mdtests/dotnet-ai/rag-pipeline/eval.yamlKey Packages
Microsoft.Extensions.AI(chat + embeddings)Microsoft.Extensions.VectorData(vector search)Microsoft.Extensions.DataIngestion(document ingestion)Validation