IntelliNode plugin
A plugin that gives an agent the IntelliNode skill and MCP server: ask any model, search, review code, tests, SQL, images and speech
- 275 ★ on the repository
- Last commit
- Apache-2.0
Broad access
Why this level
- Takes provider keys in environment variables and spends tokens against them
- The review, fix and test tools send code to external model APIs
- CodingAgent edits repository files and runs a test command
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/intellinodeDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- Takes provider keys in environment variables and spends tokens against them
- The review, fix and test tools send code to external model APIs
- CodingAgent edits repository files and runs a test command
This is third-party code. Review the repository files before installing.
What it does
IntelliNode is an npm package for JavaScript and TypeScript with one call surface for OpenAI, Anthropic, Gemini, Mistral, Cohere, NVIDIA and local models. For coding agents the repository ships a skill and an MCP server, and the plugin installs both at once. The skill teaches an agent to build chats with memory, document search, vector-store work and Gemini and Vertex AI features on top of IntelliNode. The MCP server starts with `npx -y intellinode mcp` and gives Claude Code, Codex, Cursor and VS Code the tools of every connected provider: ask a chosen model, consensus across models, web search with sources, code review, fixes, tests, component, SQL, OpenAPI and mock-data generation, plus images and speech. Provider keys go in through environment variables. The library itself has a tool-calling loop, schema-matched JSON replies, a `CodingAgent` that edits a repository until tests pass, and an MCP client for `chatbot.runTools`.
Who it is for. For JavaScript and TypeScript developers building on several AI providers who want to give Claude Code or Codex the tools of different models.
Good fit when
- You want a second opinion from another model right inside a Claude Code or Codex session
- You need a ChatGPT- or Gemini-style chat with saved conversations, memory and answers over your documents
- The app needs a vector store: Firestore, Pinecone, Qdrant, Chroma, Weaviate, Milvus, pgvector or MongoDB Atlas
- You want component, SQL, OpenAPI and mock-data generation in one tool across providers
- An agent should fix a repository until the tests pass
Not a fit when
- The project is Python: the author has a separate library, Intelli, for it
- You have no provider keys and no local model: the tools have nothing to answer with
- You need a single provider and its official SDK: a shared layer adds a dependency
- Code must not go to outside providers: review and fixes run through their APIs
Example request
Through IntelliNode MCP ask Gemini and Claude how to best design an index for document search, and compare the answersLimitations
At least one provider key is needed, for example `OPENAI_API_KEY`, `ANTHROPIC_API_KEY` or `GEMINI_API_KEY`, or a local server such as Ollama. OpenAI and Anthropic APIs are unavailable from Russia. MCP tool requests go to the chosen providers together with any code you pass in, and spend their tokens. The library requires Node.js, and the plugin starts the MCP server through `npx`. The skill is meant for JavaScript and TypeScript; the author points Python projects to a different library.
How to disable. Remove the plugin with `claude plugin uninstall intellinode@intellinode-js`, and a manually added MCP server with `claude mcp remove intellinode`. A skill installed through `npx intellinode skill install` lives in the agent's skills folder; delete it there.
FAQ
How do I install IntelliNode in Claude Code?
Run `claude plugin marketplace add intelligentnode/IntelliNode`, then `claude plugin install intellinode@intellinode-js`. The plugin installs the skill and the MCP server. The skill alone installs with `npx intellinode skill install`.
How do I connect just the IntelliNode MCP server?
For Claude Code: `claude mcp add intellinode -e OPENAI_API_KEY=sk-... -e ANTHROPIC_API_KEY=sk-ant-... -- npx -y intellinode mcp`. For Codex: `codex mcp add intellinode --env GEMINI_API_KEY=... -- npx -y intellinode mcp`.
Which models and services does IntelliNode support?
OpenAI, Anthropic, Gemini and Vertex AI, Mistral, Cohere, NVIDIA, vLLM and OpenAI-compatible services such as OpenRouter, Groq, DeepSeek and Ollama. Image, speech and video generation are there where a provider offers them.
What can CodingAgent do?
It takes a repository folder and a task, for example `Fix the failing tests in calc.js`, and with an `npm test` command repeats edits until the tests pass. The provider is chosen when the object is created.
Does a Python project need something different?
Yes. The IntelliNode skill is written for JavaScript and TypeScript, and for Python it points to the Intelli library.
Discussion
No questions or reviews yet.
Related
PluginsAn MCP server that lets Claude Code, Codex and other agents call Gemini, GPT, Grok and local models for review, debugging and consensus
A CLI that turns any MCP server, OpenAPI spec or GraphQL endpoint into commands at runtime, with no codegen
A desktop app with a local MCP server that lets ChatGPT work with files, commands, skills and other MCP servers
An MCP client for Neovim: manage, test and connect MCP servers to chat plugins right in the editor
A CLI for any MCP server: list and call tools, resources and prompts, run mock and proxy modes, guard access and manage agent configs
A CLI to run several coding agents: one config, a session console, local memory and a set of engineering skills
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| IntelliNode (this entry) | no | 275 | ||
| just-prompt | no | key is optional | 739 | an MCP server that fans one prompt out to several models, with no library or skill |
| Oracle | no | 4,060 | a CLI that packs a prompt and project files and sends them to another model for a second opinion |
README in short
The README presents IntelliNode as one prompt and evaluation layer for any large model and lists what is new in 3.0 and 3.1. It then gives JavaScript call examples: Gen functions, chatbots for several providers, Assistant and vector stores, followed by sections on the skill and the MCP server for coding agents.