Deploy and query retrieval pipelines as a single RoutIR service.
npx skills add https://github.com/hltcoe/routir --skill routir
Please help me install this Agent Skill. Skill: routir Source: https://github.com/hltcoe/routir/tree/main Command: npx skills add https://github.com/hltcoe/routir --skill routir
RoutIR removes the friction of serving retrieval models by giving you one service that can handle search, reranking, query expansion, score fusion, and document access behind a consistent API.
Ask RoutIR to start from your JSON config and then query the exposed services with the client or pipeline DSL.
High-intent search queries and answers about installing and using this skill.
Deploy RAG retrieval pipelines as a service by providing a JSON configuration file to define dense search, sparse retrieval, reranking, and fusion stages. The service exposes these stages behind unified HTTP or gRPC APIs for client access.
Combine dense and sparse retrieval by using a pipeline DSL to chain search, reranking, query expansion, and score fusion stages. This composable approach builds hybrid multi-stage RAG retrieval flows within a single service.
You can serve single-view or multi-view document collections that include both text and bytes content. This collection support enables multimodal reranking and view-aware content routing through the configured HTTP or gRPC clients.
The retrieval service supports both gRPC and REST APIs. You can query exposed dense search, sparse retrieval, and reranking pipelines using clients that handle batching, caching, and authentication.
Add authentication and caching to retrieval pipelines by configuring operational deployment features within your JSON config. This production setup also supports configurable batching and remote service relays for RAG workflows.
Use remote service relays in RAG workflows when you need to distribute or proxy retrieval tasks across external services. Relays allow your pipeline to integrate remote dense search, sparse retrieval, or reranking engines behind a consistent API.
anukkrit149Build Retrieval-Augmented Generation systems grounding LLM answers in vector databases.
hellonishSearch YouTube videos and extract transcripts with channel and duration metadata.
zhouboyu-xrealSearch and retrieve academic papers from arXiv via its REST API.
pomazanbohdanDeploy a Rust-native RAG pipeline with Candle embeddings and hybrid retrieval.
