About LeadMagic
LeadMagic's public ChatGPT and Codex package connects to its hosted service for B2B account research, contact enrichment, table work and connected CRM inspection. It is intended for users working with LeadMagic; available tools depend on workspace permissions, and paid table runs require scope, a credit quote and human approval.
Key features
- 01B2B account research
- 02Contact enrichment
- 03Table inspection and scoped cleanup
- 04Connected CRM inspection
- 05Approval flow for paid table runs
Use cases
- 01Research a B2B account
- 02Enrich contact records
- 03Inspect and clean a LeadMagic table
- 04Review connected CRM information
MCP support
The package declares LeadMagic's hosted Streamable HTTP MCP endpoint. Clients authenticate through the service's OAuth flow; this repository contains connection metadata, not the server implementation.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://mcp.leadmagic.io/mcpclaude mcp add --transport http io-leadmagic-server https://mcp.leadmagic.io/mcp{"mcpServers":{"io-leadmagic-server":{"type":"http","url":"https://mcp.leadmagic.io/mcp"}}}Open in Cursor{"servers":{"io-leadmagic-server":{"type":"http","url":"https://mcp.leadmagic.io/mcp"}}}Open in VS Code{"mcpServers":{"io-leadmagic-server":{"type":"http","url":"https://mcp.leadmagic.io/mcp"}}}{"mcpServers":{"io-leadmagic-server":{"serverUrl":"https://mcp.leadmagic.io/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMELeadMagic for ChatGPT
The public LeadMagic package for ChatGPT and Codex, published by LeadMagic Corporation. It contains directory metadata, the hosted MCP connection at https://mcp.leadmagic.io/mcp, declared listing images, and the LeadMagic Tables skill. Authentication is configured in the OpenAI submission portal.
LeadMagic supports B2B account research, contact enrichment, table inspection and scoped cleanup, and connected CRM inspection. Available tools depend on the connected workspace's permissions. Paid table runs require a scope and credit quote followed by human approval. Outbound CRM and sequencer delivery is handled in the LeadMagic app.
Build
Requires Node.js 24.18+ and zip. Authored tooling uses TypeScript with Node's native type stripping; validation also runs the strict TypeScript compiler and package boundary tests.
npm ci
npm run validate
npm run package -- --out=/tmp/leadmagic-openai-plugin.zip
Support: LeadMagic Support · Privacy · Terms
The validator requires the public author name and developer name to be exactly LeadMagic Corporation. It checks the stable plugin identity, listing limits, review tool names, and one hosted HTTPS MCP without embedded credentials, headers, or an authentication override. It rejects app references, lifecycle hooks, unsafe paths, symlinked package files, and obsolete paid-run instructions. This package uses noninterlaced PNG images, decoded with pinned pngjs after checking file size and dimensions. Declared screenshots must be distinct PNG assets, one per starter prompt, each 706 pixels wide and 400–860 pixels tall, at most 5 MiB.
The upload ZIP contains only the plugin manifest, MCP connection metadata, declared images, and the Tables skill. Packaging creates a fresh archive, so previous ZIP contents cannot survive. Registry metadata, build tools, repository files, and undeclared assets are excluded. Write the ZIP outside this repository. Keep reviewer credentials and sign-in instructions in the secure OpenAI review form; OpenAI rejects them in ZIP metadata.
To check review tool names against a saved MCP tools/list response or catalog:
npm run validate -- --tools=/absolute/path/tools-list.json
A local catalog verifies tool names only. It is not a fresh authenticated production scan.
Submit an update
Use the existing LeadMagic entry in the OpenAI submission portal and the With MCP flow. Preserve the package identity and existing authentication configuration, verify domain ownership, and scan the production endpoint. Screenshots require custom UI reported by that scan.
Run all five positive and three negative review cases using the review account and provide an accessible demo recording that demonstrates the current package. Local validation does not verify live tool behavior or complete directory review.
Package rules: OpenAI submission guide · Submission errors · Plugin guidelines.
Official MCP Registry
server.json publishes LeadMagic's hosted MCP endpoint to the Official MCP Registry. It uses the io.leadmagic domain namespace and a Streamable HTTP connection; clients authenticate through the service's OAuth flow. This repository contains public connection and listing metadata only, not the hosted server implementation.
Validate the listing before publishing with mcp-publisher validate server.json.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.