superrag-build

superrag-build is a skill for Claude Code, Codex from Aryanacoder/superragskills. It costs 131 tokens per session (1,857 once invoked), scanned A, original, MIT.

An end-to-end guide for building retrieval-augmented generation systems. These systems retrieve relevant information before an AI writes an answer, and may use documents, databases, images, graphs, or structured data.

In plain words
What is it for?
Planning, building, evaluating, and deploying document copilots, source-grounded chatbots, ingestion pipelines, hybrid or agentic search systems, and local or cloud RAG applications.
Why use it?
RAG projects involve choices about data, search quality, citations, privacy, cost, and deployment. The guide structures those choices before implementation and testing.

Skill for Claude CodeCodex

Written for Claude Code and Codex: shipped in a Claude Code plugin, but also agents/openai.yaml present.

Part of the superrag-build plugin — 1 skill shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/aryanacoder/superragskills/superrag-build
Any agent
npx skills add Aryanacoder/superragskills --skill superrag-build
Clone the repo
git clone --depth 1 https://github.com/Aryanacoder/superragskills

Made for: Claude Code, Codex.

Or install superrag-build, the plugin that ships this one along with the rest of its 1 skill.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for superrag-build

README.md
[![agentmods](https://agentmods.dev/badge/skills/aryanacoder/superragskills/superrag-build.svg)](https://agentmods.dev/skills/aryanacoder/superragskills/superrag-build)
Your own site
<a href="https://agentmods.dev/skills/aryanacoder/superragskills/superrag-build"><img src="https://agentmods.dev/badge/skills/aryanacoder/superragskills/superrag-build.svg" alt="Measured on agentmods" height="20"></a>
Per session 131 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,857 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00131 $0.01857
Opus 5 $0.00066 $0.00928
Sonnet 5 $0.00026 $0.00371
Haiku 4.5 $0.00013 $0.00186

Measured 5d ago against content hash 8bb332cbdfdc, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

superrag-build scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/superrag-build/SKILL.md · 131 lines

How it starts

The opening of the file, as written. The whole thing — 131 lines — stays where its author put it; the contents beside it link to each section on GitHub.

superRAG-build

Mission

Be the RAG architect, interviewer, implementation planner, and deployment guide. Turn an unclear RAG idea into the simplest architecture that can satisfy the user's data, risk, quality, budget, and deployment constraints.

Do not jump straight to code for broad RAG requests. First classify the RAG type, choose the stack, define the evaluation path, and identify deployment constraints. If the user asks for a tiny prototype, ask only the missing high-risk questions and build a narrow vertical slice.

First Response

Ask the first batch of questions, then adapt:

  1. What job should this RAG app help users complete?
  2. Who are the users, and what is the cost of a wrong answer?
  3. What data exists: file types, volume, languages, scans, images, tables, databases, websites, code, APIs?
  4. Does every answer need citations, page previews, exact quotes, source downloads, or audit logs?
  5. Is this local/offline, cloud, hybrid, multi-tenant, or enterprise/private?
  6. Preferred stack: AWS, Azure, OpenAI-hosted, open source, LangChain, LlamaIndex, custom, no preference?
  7. What is the first demo query that must work?

Then say which RAG families are likely candidates and what you need next. Ask questions in batches; do not dump the whole questionnaire unless the user asks for a worksheet. Use references/question-bank.md for exhaustive discovery.

Architecture Selection

Use references/rag-types-decision-matrix.md before choosing tools.

Default choices:

  • Naive/basic RAG: small corpus, low risk, natural-language lookup, quick prototype.
  • Hybrid RAG: exact terms, IDs, part numbers, policy clauses, support tickets, code, procedures, or mixed vague/specific queries.
  • Reranked RAG: quality matters and first-stage retrieval returns too many candidates.
  • Metadata-filtered RAG: tenants, permissions, product/version filters, regions, dates, languages, document status.
  • Hierarchical/parent-child RAG: long PDFs, books, policy docs, manuals, multi-section reports.
  • Agentic RAG: multi-hop questions, query planning, iterative retrieval, ambiguity, tool use, cross-document synthesis.
  • Corrective/Self-RAG: retrieval may be weak and the system must evaluate, retry, or refuse.
  • GraphRAG: entity relationships, communities, investigations, knowledge discovery, cross-document relationship queries.
  • Multimodal RAG: images, diagrams, scans, charts, screenshots, video frames, CAD/flowcharts, visual tables.
  • Structured-data RAG: SQL/BI/metrics/doc hybrids where the answer needs database queries plus text evidence.
  • Streaming/real-time RAG: fast-changing data, events, logs, tickets, CDC, queues.
  • Local/offline RAG: air-gapped, privacy, edge machines, field laptops, classified or regulated data.

Read the full file on GitHub · 131 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago First seen · 131 lines · 131 tokens per session scan A 8bb332cbdfdc

Subscribe to this mod's changes

superrag-build is a skill published in the GitHub repository Aryanacoder/superragskills (3 stars, last pushed 2mo ago), licensed MIT. It adds 131 tokens to every session and 1,857 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

embeddings

Vector embeddings with HNSW indexing, sql.js persistence, and hyperbolic support. 75x faster with agentic-flow integration. Use when: semantic search, pattern matching, similarity queries, knowledge retrieval. Skip when: exact text matching, simple lookups, no semantic understanding needed.

ruvnet/ruflo · 62 tokens

agent-platform-rag-engine-management

Manage and query Agent Platform RAG Engine Corpora and retrieve grounded contexts using the Google GenAI SDK. Use when listing RAG corpora or files, inspecting a corpus, retrieving contexts, or generating content grounded in a RAG corpus. Do not use for standard database queries (use SQL/Spanner skills), Google…

google/skills · 85 tokens

llm-app-patterns

Production-ready patterns for building LLM applications. Covers RAG pipelines, agent architectures, prompt IDEs, and LLMOps monitoring. Use when designing AI applications, implementing RAG, building agents, or setting up LLM observability.

davila7/claude-code-templates · 54 tokens

9router-embeddings

Generate vector embeddings via 9Router /v1/embeddings using OpenAI / Gemini / Mistral / Voyage / Nvidia / GitHub embedding models for RAG, semantic search, similarity. Use when the user wants embeddings, vectors, RAG, semantic search, or to embed text.

decolua/9router · 66 tokens

azure-search-documents-dotnet

Azure AI Search SDK for .NET (Azure.Search.Documents). Use for building search applications with full-text, vector, semantic, and hybrid search. Covers SearchClient (queries, document CRUD), SearchIndexClient (index management), and SearchIndexerClient (indexers, skillsets). Triggers: "Azure Search .NET"…

microsoft/skills · 102 tokens

similarity-search-patterns

Implement efficient similarity search with vector databases. Use when building semantic search, implementing nearest neighbor queries, or optimizing retrieval performance.

foryourhealth111-pixel/Vibe-Skills · 30 tokens