rag-manager

Manage RAG indices via shell scripts against the BTDP RAG API.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/smorand/claude-config --skill rag-manager
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: rag-manager
Source: https://github.com/smorand/claude-config/tree/main/skills/rag-manager
Command: npx skills add https://github.com/smorand/claude-config --skill rag-manager

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq, curl, and includes scripts (resource) components.

What problem does it solve?

This Skill centralizes RAG index operations (list, search, create, delete) and provides reliable shell scripts for interacting with the BTDP RAG API.

Core Features & Use Cases

  • List & summarize indices: Retrieve all configured RAGs with details.
  • Semantic search: Query indices using natural-language-like prompts.
  • Create & delete indices: Manage embeddings and associated resources.
  • API-driven workflows: Integrate with BTDP RAG API for automation.

Quick Start

Use the provided rag_list.sh, rag_search.sh, rag_create.sh, and rag_delete.sh to manage indices.

Frequently Asked Questions about rag-manager

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I manage RAG indices for semantic search in BTDP?

Semantic RAG index management centralizes list, search, create, and delete operations against the BTDP RAG API. Use the provided shell scripts (rag_list.sh, rag_search.sh, rag_create.sh, rag_delete.sh) with OAuth authentication to manage embeddings and enable natural-language queries across C1 and C2 datasets.

Can I query RAG indices using natural language prompts?

Yes. Semantic search queries RAG indices using natural-language-like prompts via rag_search.sh against the BTDP RAG API. Results return structured JSON with matching embeddings and metadata from your configured indices.

What authentication does the RAG API require?

The BTDP RAG API authenticates using an OAuth token stored at ~/.gcp/access_token. The shell scripts automatically read and apply this token to API calls, enforcing secure access to C1 and C2 data handling policies.

How do I create and delete RAG indices programmatically?

Use rag_create.sh to provision new indices with embeddings and rag_delete.sh to remove them. Both scripts handle API communication, error handling, and return structured JSON results for automation in BTDP workflows.

Does this work with BigQuery datasets?

Yes. RAG index management applies to L'Oréal's BTDP infrastructure, which integrates with BigQuery for C1 and C2 dataset operations. Indices enable semantic search across your BigQuery data through the BTDP RAG API.

What are the limitations of shell-script-based RAG operations?

Operations depend on jq and curl availability, OAuth token accessibility at ~/.gcp/access_token, and network connectivity to the BTDP RAG API endpoint. Large-scale index operations may require batching; refer to API response limits in BTDP documentation.