Lumi App Finder MCP
Gives AI assistants a read-only tool to match a user's task or buyer need to verified Lumi Studio iOS apps across 29 apps and all 50 Apple locales, returning localized context, purchase model, guide, and direct App Store links. Supports interactive localized cards via the MCP Apps extension and includes a bundled offline snapshot without analytics.
README
Lumi App Finder MCP
Lumi App Finder gives AI assistants one read-only tool for matching a user's task or buyer need to a verified live Lumi Studio iOS app.
It covers 29 apps × all 50 Apple locales. Every result includes editorially localized context, the purchase model, a detailed guide, and a direct App Store link for the matching storefront.
Hosts that support the stable MCP Apps extension render the matches as interactive, localized cards with one-click App Store and guide actions. Other hosts receive the same complete text and structured-data fallback.
First-party disclosure: Lumi Studio develops every listed app. Results are transparent publisher-authored text matches, not measured search volume, independent rankings, reviews, or user endorsements.
Agent Skill
AI hosts that support the open Agent Skills specification can install the same 50-locale, first-party catalog as an offline, progressively loaded skill:
gh skill install alice51849/lumi-mcp lumi-app-finder@v1.1.3 --scope user
GitHub CLI 2.90 or newer installs to GitHub Copilot by default. Add
--agent claude-code, --agent cursor, --agent codex, or
--agent gemini-cli for another supported host. The installed skill reads only
the requested locale's 29-record snapshot and requires no account, API key,
network request, or executable script at runtime.
The version-pinned skill also works with the Vercel Skills CLI and its supported agents:
npx -y skills@1.5.19 add https://github.com/alice51849/lumi-mcp/tree/v1.1.3/skills/lumi-app-finder --skill lumi-app-finder -g -y
The third-party installer reports anonymous installation telemetry by default under the skills.sh CLI policy. The installed Lumi App Finder skill itself remains offline and contains no analytics.
The public skill is discoverable on its
skills.sh directory page,
which reads the same first-party SKILL.md from this repository.
Lumi App Finder is also listed in the independently maintained
Awesome Skills directory,
which links to the immutable v1.1.3 skill files.
Tool
find_ios_apps
Inputs:
query— task, app name, or buyer need in any supported language.locale— one of Apple's 50 supported locale codes; defaults toen-US.limit— 1–10 matches; defaults to 5.
The server reads the current public catalog when online and falls back to the bundled 1,450-record snapshot. User query text never leaves the local MCP process.
Install
Find io.github.alice51849/lumi-app-finder in clients or registries that use
the official MCP Registry, or use a version-pinned installer:
The VS Code and Cursor links run the zero-dependency server from the pinned
public v1.1.3 GitHub release through npx; Node.js 20 or newer is required.
The same pinned configuration can be added manually:
{
"servers": {
"lumi-app-finder": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"https://github.com/alice51849/lumi-mcp/releases/download/v1.1.3/lumi-app-finder-npx.tgz"
]
}
}
}
The MCPB uses Node's stdio transport and needs no account, API key, external UI runtime, or manual configuration.
For Cline, use the exact configuration and verification steps in llms-install.md.
50-locale coverage
ar-SA, bn-BD, ca, cs, da, de-DE, el, en-AU, en-CA,
en-GB, en-US, es-ES, es-MX, fi, fr-CA, fr-FR, gu-IN, he,
hi, hr, hu, id, it, ja, kn-IN, ko, ml-IN, mr-IN, ms,
nl-NL, no, or-IN, pa-IN, pl, pt-BR, pt-PT, ro, ru, sk,
sl-SI, sv, ta-IN, te-IN, th, tr, uk, ur-PK, vi,
zh-Hans, zh-Hant.
MCPB display metadata and the catalog output both use these localized resources.
Development
npm ci
npm run build:ui
npm run build:skill
npm test
npm run validate
npm run pack:mcpb
Refresh the bundled catalog and 50 locale resources from the public
ios-app-guide source:
npm run sync:catalog
Version tags publish the bundle as a GitHub release and register it through GitHub OIDC, without long-lived registry credentials.
Privacy
See PRIVACY.md. The tool has no analytics or tracking, and never sends user queries to the catalog host.
Security
Report vulnerabilities privately through the security policy, not through a public issue.
License
MIT. Third-party language-data notices are in THIRD_PARTY_NOTICES.txt; bundled MCP App library licenses are in MCP_APP_NOTICES.txt.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.