RAG Eval

Scores retrieval quality on a fixed question set.

Single skill Verified 10 Nov researchragevaluation

Use it for

Answering whether a retrieval change made the system better

Skip it if

Building the pipeline — it evaluates, it does not index

What it is

Runs a fixed set of questions against a retrieval pipeline and reports two numbers: whether the expected document appeared, and whether the answer cited it. Both matter, because a system that retrieves the right document and cites a different one is worse than it looks in a demo.

Install

Clone the repository, then copy the skill folder into your agent's skills directory — ~/.claude/skills/ for Claude Code. Skills are plain files: no build step, no daemon.

Terminal
npx skills add rag-eval

How to judge it

Read the SKILL.md before you install. A good skill states when it applies, what it must not do, and how to verify the result. A skill that only lists adjectives will change the agent's tone without improving its output — skip those.

Other research skills

  • Competitor Scan — Collects what changed on a competitor's site since the last scan.
  • SQL Analyst — Turns a business question into one reviewed query.
  • Data Cleaning — Profiles a messy CSV and reports what would break a join.

Tools it pairs with

All tools

Servers it runs on

All servers
  • 01
    Telegram MCPCommunity

    Reads channels and groups through the Bot API: recent messages, search, and posting. The distinction that matters is bot token versus user session — a bot sees what …

    #chat#monitoring#research
    Pythonstdio
  • 02
    YouTube MCPCommunity

    Wraps the YouTube Data API for search, channel and video metadata, and captions where available. The transcript path is the useful one for research; anything that re…

    #video#research#transcripts
    Pythonstdio
  • 03
    Reddit MCPCommunity

    Searches subreddits and reads threads with comments through the Reddit API. It is a reasonable research input and a poor automation target: the platform's rules and …

    #research#social#sentiment
    TypeScriptstdio