Skip to content
Metatext Decide — typed decisions for your agents, with a probability on every answerGet early access
Metatext
Sign Up

apple-rag-mcp

MCP server over Apple developer docs and WWDC transcripts that uses hybrid search, then Jev relevance scores to rank results.

MCP serverproductionMIT
View on GitHub ↗Live demo ↗

Overview

apple-rag-mcp gives AI agents one lean MCP tool for Apple developer documentation and WWDC video transcripts. Hybrid retrieval combines keyword and semantic search to gather candidates, and Jev then scores each candidate against the query's API, platform and version requirements so the best matches are returned. A backup reranker takes over if the Jev call fails.

Responses are kept small to avoid filling agent context. A hosted service at apple-rag.com provides an already collected, cleaned and indexed knowledge base, with embeddings, Jev ranking and the backup reranker included. The source is public for inspection and self-hosting, which means running your own database and index refresh.

What the decision model decides

TypeQuestionAnswers
ScoreHow relevant is this Apple doc or transcript to the query's API, platform and version?(inferred)—

How it works

  1. Agent asks a question via MCP
  2. Hybrid keyword and semantic retrieval gathers candidates
  3. Jev scores candidates for relevance
  4. Return top results; fall back to a backup reranker on failure

Strengths

  • Lean responses save agent context
  • Fallback reranker for resilience
  • Hosted option with model calls included

Caveats

  • Self-hosting requires running a database and indexing
  • Hosted service is a dependency

Getting started

Connect your MCP client to the hosted service at https://apple-rag.com, or self-host from source.

At a glance

Type
MCP server
Status
production
Language
TypeScript
Runtime
MCP server (hosted or self-hosted)
Has code
yes
Has tests
yes
License
MIT

Use cases and integrations

How curated lists describe it

Frequently asked questions

What does Jev do?

It scores retrieved Apple sources for relevance so they can be ranked.

What if Jev fails?

A backup reranker takes over.

Do I need my own keys for the hosted service?

The README says retrieval model calls are included in the hosted service.

Related projects

  • hippo-memory

    Local long-term memory for coding agents with an optional Jev reranker that orders recalled memories.

    ProjectproductionTypeScript
  • BGTS Context Engine

    Deterministic code-graph context for coding agents on Postgres, AGE and pgvector, with optional Jev or local re-selection of symbols.

    MCP serverbetaPython
  • jev-rerank

    Agent skill and Python library for Claude Code and Codex that reranks candidate lists with one Jev yes/no relevance question each.

    ProjectbetaPython
  • jevex

    Single MCP tool that indexes a repo and has Jev rank a shortlist so a coding agent reads the right files and lines instead of grepping.

    MCP serverbetaPython
  • JevGrep (Arifette)

    CLI and MCP server where Jev scores code fragments for a behaviour question and returns exact excerpts with line numbers.

    MCP serverbetaTypeScript
  • kody

    Multi-user MCP 'agents' cloud' on Cloudflare Workers with an optional search stage that reranks candidates with Workers AI Jev scores.

    MCP serverbetaTypeScript