Razaurrehman

Retriva — AI skill for Claude Code

AI community

Self-hosted, production-ready RAG chatbot agent — upload documents, get accurate AI answers with citations, agentic tool/API calling, and multi-turn memory.

How to install Retriva

This entry records only its repository, not the path inside it, so there is no exact command to give. Open Razaurrehman/retriva and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Retriva does

Self-hosted, production-ready RAG chatbot agent — upload documents, get accurate AI answers with citations, agentic tool/API calling, and multi-turn memory. Built with Claude, Voyage AI & Qdrant. Embed into e-commerce, CRM, or support apps.

Alternatives in AI

WeKnora — Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, an 26.9k ★
Welcome — AI Research Skills — You now have access to 86 production-ready skills covering the entire AI research lifecycle: literature survey 5.4k ★
Contributing To Claude Skills Library — Thank you for your interest in contributing to the Claude Skills Library 5.3k ★

README

Retriva

**Retriva** is a self-hosted, production-ready **RAG chatbot agent** — an open-source AI assistant you can train on your own documents in minutes and embed into any application. Built for teams who want an accurate, controllable **knowledge-base chatbot** without handing their data to a black-box SaaS.

Upload PDFs, Word docs, spreadsheets, or plain text; ask questions; get clean, structured JSON answers with real citations back to the source. Retriva also does **agentic tool/API calling** out of the box — it ships with a working example (order-status lookup) so you can see exactly how to wire in your own live systems (order management, ticketing, a CRM, inventory, anything with an API).

**Why Retriva:**

🔍 Accurate retrieval, not guesswork — vector search over your own documents, with citations on every answer
🛠️ Agentic tool use — the LLM decides when to search your knowledge base vs. call a live API, the same way modern AI agents work
💬 Real conversation memory — multi-turn context with automatic summarization, not a stateless Q&A box
📦 Train by uploading, not fine-tuning — drop in a document, it's searchable in seconds, no ML expertise required
🔌 Embed anywhere — a clean REST API built for e-commerce, CRM, customer support, internal tools, or any app that needs an AI assistant grounded in real data
🔒 Self-hosted and private — your documents and conversations never leave your own infrastructure
⚡ Production-ready from day one — rate limiting, fail-fast config validation, structured error handling, and a real test suite, not a weekend prototype

Powered by **Claude** (Anthropic), **Voyage AI** embeddings, and **Qdrant** vector search — the same class of stack behind today's leading AI agent products, running entirely under your control.