M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
N

NeuroLink

Community
Framework / SDK
The TypeScript pipe layer for the AI nervous system.

This is a framework for building MCP servers or agents, not a server you connect a client to.

How to install ↓ Compare GitHub source repository ↗
Category
Dev Tools #295 of 438
Stars
★ 157 Very popular
Runtime
Node.js
Credentials
API key / credential required
License
MIT
Last commit
Tools
6
46FMRS · D

NeuroLink is a broad Juspay AI integration platform in which MCP is one connectivity and tool-management capability. Because the source lacks a cached server manifest, it should not be presented as an independently verified MCP server with a complete tool list or transport declaration.

Strongest · Documentation 11/20 Weakest · Reliability 8/20

Reliability
8/20
Security and permissions
8/20
Maintenance
10/20
Documentation
11/20
Setup experience
9/20
Why each score
Reliability 8/20
The static evidence is limited to the README and metadata; the server manifest is not cached, and there is no verifiable server launch configuration, MCP initialize/tool-list handshake implementation, or committed test coverage for key paths. The README claims built-in tools, multiple transports, and extensive functionality, but those behaviors cannot be verified from the supplied material, so points are deducted; no explicit breakage is shown, leaving only limited plausibility credit.
Security and permissions 8/20
The README describes credentials via environment variables or per-request input, HITL, auditing, and input sanitization, but none of these controls is supported by source or test evidence here. Built-in writeFile, external stdio command execution, remote HTTP/SSE/WebSocket connections, and GitHub/database side effects broaden privileges and data boundaries; confirmation flow, default permissions, sandboxing, network scope, and sensitive-data handling remain unclear, so points are deducted. The supplied material does not establish malware, credential theft, or an unconfirmed destructive default, so no red line is triggered.
Maintenance 10/20
The repository is not archived, explicitly identifies the MIT license, and attributes the project to Juspay; the README lists multiple version updates, suggesting possible maintenance activity. However, the supplied material contains no commit history, verified release records, dependency-update evidence, maintainer response process, vulnerability disclosure policy, or security-response channel. Stars and issue counts add no points, so the score remains moderate.
Documentation 11/20
The README provides installation, SDK/CLI examples, provider setup, MCP transport descriptions, a broad feature overview, and some troubleshooting, giving it substantial coverage. However, the server manifest is absent, and there are no reviewable MCP tool schemas, parameter/return definitions, permission boundaries, limits, costs, defaults, or complete mainstream-client configuration. The README also varies between 24+, 30+, and 100+ scope claims, while many capabilities cannot be verified from the supplied source, so points are deducted.
Setup experience 9/20
npm/pnpm/yarn installation, CLI invocation, and environment-variable configuration are clearly shown, but the material primarily documents SDK/CLI usage rather than a complete configuration for connecting this server from a mainstream MCP client. At least one external model provider must be selected and configured, while the MCP manifest and server launch command are missing; the multiple providers, credentials, transports, and optional infrastructure make the first working path fairly complex, so the score is below the upper-middle range.

Static review · not runListed 2026-08-14

Read the FMRS scoring method →

Fit and risk

What it can accessReads local filesWrites / deletes local filesUses the network

Best for

  • TypeScript teams needing a unified API across multiple AI providers
  • Applications requiring MCP integration, RAG, memory, or provider failover
  • Developers who want both an SDK and a professional CLI

Not for

  • Users seeking a standalone MCP server with a verified tool manifest and transport
  • Users unwilling to configure AI provider credentials
  • Simple MCP use cases that do not need a TypeScript SDK, CLI, or multi-provider features

Required permissions

  • API credentials for configured AI providers
  • Appropriate local file read permissions when processing files
  • File write permissions when using the built-in writeFile tool
  • Tokens, environment variables, or authorization required by connected external MCP servers
  • ffmpeg or bundled ffmpeg-static may be required for video processing

Risks and side effects

  • Prompts, files, or audio may be sent to the selected AI provider; data handling depends on provider configuration
  • writeFile and external MCP tools may modify data or trigger external actions
  • Invalid credentials, rate limits, or provider outages can cause requests to fail
  • Tokens and permission scopes for external MCP servers affect the security boundary
  • Audio, video, and large-context processing may require additional dependencies or resources

Setup

Before you start

Runtime：Node.js

OPENAI_API_KEY optionalsecret API key from platform.openai.com for OpenAI GPT models.
ANTHROPIC_API_KEY optionalsecret API key from console.anthropic.com for Anthropic Claude models.
GOOGLE_AI_API_KEY optionalsecret API key from aistudio.google.com (free tier) for Gemini models.
GITHUB_TOKEN optionalsecret GitHub personal access token generated in GitHub settings, used when connecting the GitHub MCP server.
Other optional settings (1)
REDIS_URL optional Redis connection URL for distributed conversation memory; set it after starting Redis, e.g. via Docker.

Install @juspay/neurolink with npm, pnpm, or yarn, or run the CLI with npx. Configure at least one AI provider API key through the setup wizard or environment variables. When connecting external MCP servers, configure the required command or URL, transport, and credentials for each server.

Check that it works

After installing @juspay/neurolink and configuring at least one provider API key, run npx @juspay/neurolink status to check provider health, or confirm built-in tools like getCurrentTime, readFile, writeFile, listDirectory, and calculateMath appear in the client's tool list.

Troubleshooting

  1. Run setup or confirm that at least one provider API key is configured
  2. Check that provider and model names match the configured credentials
  3. For rate limits or outages, configure multiple providers or use fallback/modelChain
  4. For video failures, check that ffmpeg is available
  5. For Redis memory, verify REDIS_URL, connection settings, and Redis availability
  6. For external MCP connections, verify the command, URL, transport, headers, environment variables, and retry settings

Things to try

Once connected, you can ask your AI assistant things like:

  • What time is it right now?
  • List all files in the current directory
  • Calculate (128 * 47) + 1024 for me
  • Write the meeting summary into notes.txt

Tools 6

getCurrentTime read-only
Provides access to the current time.
readFile read-only
Reads file contents.
writeFile writes
Writes file contents.
listDirectory read-only
Lists directory contents.
calculateMath read-only
Performs mathematical operations.
websearchGrounding read-only
Provides Google Vertex web search and requires credentials.

Use cases

Switching between AI providers without rewriting application code
Building AI applications with streaming, RAG, memory, and multimodal file processing
Connecting external MCP servers such as GitHub, Slack, and PostgreSQL
Running generation, evaluation, RAG, and interactive sessions through the CLI
Building TTS, STT, or realtime voice applications

Overview

NeuroLink is a TypeScript-first AI integration platform extracted from Juspay production systems. It provides one API for 30+ AI providers and 100+ models, with streaming, MCP server connectivity, voice, RAG, memory, multimodal file processing, provider failover, and a CLI. The source does not provide a standalone MCP server manifest, so its independently exposed MCP tools and transport cannot be verified.

Similar servers

Context7 80 · B

Upstash's official server providing up-to-date third-party library docs for AI coding assistants

★ 63.3k · Tools 2 Compare with this →
Compare with other MCP servers

Source revision 45557b3bef6f Data synced 2026-10-19 Read the FMRS scoring method