Apple Notes MCP vs mcp-agent

A side-by-side look at pricing, capabilities, pros, cons, and our editorial scores.

	Apple Notes MCP MCP Servers	mcp-agent MCP Servers
Tagline	Let Claude read your local Apple Notes over the Model Context Protocol.	Python framework for building composable AI agents on the Model Context Protocol
Category	MCP Servers	MCP Servers
Pricing	Free· Free / open-source (MIT). No hosted service; runs locally on your Mac.	Free· Free and open source (Apache 2.0). LastMile AI offers an optional managed cloud/deployment tier (Beta) with usage-based pricing not publicly listed at time of writing.
Model	—	Provider-agnostic — works with OpenAI (GPT-4o family), Anthropic (Claude 3.5/3.7), Google (Gemini), Azure OpenAI, and AWS Bedrock
Editorial score	—	—
Use cases	Personal knowledge retrieval from Apple NotesSearching decade-old meeting notes during a Claude chatPulling travel or recipe notes into a planning conversationSummarising a specific note by titleBuilding a local MCP toolchain alongside filesystem and browser serversReference implementation for writing your own macOS SQLite-backed MCP server	MCP-based deep research agentOrchestrator-worker document processingRouter-based customer support triageEvaluator-optimizer content refinement loopsMulti-agent swarm for code reviewDurable long-running research workflows on TemporalExposing an internal agent as an MCP server for Claude DesktopParallel map-reduce over large document setsIntent classification and hand-off between specialist agents
Pros	Genuinely local: reads the on-disk SQLite database directly, so notes never leave your Mac Trivial install via uv/uvx and a short claude_desktop_config.json block Exposes the three operations that matter most — list, read, search — with a clean MCP surface MIT-licensed Python, small enough to audit or fork in an afternoon Works with any MCP client, not just Claude Desktop, so it composes with other servers No API keys, no subscriptions, no accounts — zero ongoing cost	MCP-native from the ground up — any MCP server (filesystem, GitHub, Slack, browser, custom) is immediately usable without writing adapters Ships composable implementations of Anthropic's canonical agent patterns (router, orchestrator-worker, evaluator-optimizer, swarm, deep research) Durable execution via Temporal is opt-in — the same agent code runs on asyncio locally and pauses/resumes on Temporal in production Multi-provider: OpenAI, Anthropic, Google, Azure, and AWS Bedrock supported behind a common interface Agents can themselves be exposed as MCP servers, making them callable from Claude Desktop, Cursor, or any MCP-aware client Built-in OpenTelemetry tracing and token accounting for real observability, not just print-debugging Apache 2.0 licensed, active repo (8k+ stars) with regular releases and a healthy examples directory
Cons	Read-only: cannot create, edit, or append notes from within Claude Skips password-protected notes entirely (ZISPASSWORDPROTECTED is unhandled) No attachment content, no checklist state, no pinned-note filtering, no iCloud sync awareness macOS-only by design, and requires granting Full Disk Access to the runner Repository is archived by the maintainer — bug fixes and new features are unlikely without a fork Search is basic keyword matching against the SQLite text, not semantic retrieval	Python-only — no first-class TypeScript/JavaScript port for teams standardized on Node You still have to run and secure the underlying MCP servers yourself; the framework does not host them for you Durable execution requires operating a Temporal cluster (self-hosted or Temporal Cloud), which is meaningful infra overhead The hosted mcp-agent Cloud deployment product is still labeled Beta, so production-grade managed hosting is not fully mature Fewer prebuilt integrations and less community tutorial content than LangChain/LangGraph, so you will read source more often Rapidly evolving API surface — minor releases still land breaking changes as MCP itself matures
Website	github.com	github.com

Pick Apple Notes MCP if

✅ Genuinely local: reads the on-disk SQLite database directly, so notes never leave your Mac
✅ Trivial install via uv/uvx and a short claude_desktop_config.json block
✅ Exposes the three operations that matter most — list, read, search — with a clean MCP surface
✅ MIT-licensed Python, small enough to audit or fork in an afternoon

Pick mcp-agent if

✅ MCP-native from the ground up — any MCP server (filesystem, GitHub, Slack, browser, custom) is immediately usable without writing adapters
✅ Ships composable implementations of Anthropic's canonical agent patterns (router, orchestrator-worker, evaluator-optimizer, swarm, deep research)
✅ Durable execution via Temporal is opt-in — the same agent code runs on asyncio locally and pauses/resumes on Temporal in production
✅ Multi-provider: OpenAI, Anthropic, Google, Azure, and AWS Bedrock supported behind a common interface

Compare a different pair →