Rinda Ai

RINDA AI — B2B Export Sales Automation

A Claude Code plugin marketplace for B2B export sales automation — find overseas buyers, enrich contacts, run email campaigns, manage replies, and track performance.

한국어 README 보기 (Korean)

---

Quick Start

# 1. Install the marketplace
/plugin marketplace add FINGU-GRINDA/claude-rinda-plugin

# 2. Connect your RINDA account
/rinda-ai:rinda connect my account

# 3. Start using it
/rinda-ai:rinda find cosmetics importers in the US

The CLI binary installs automatically on first session.

---

What It Does

Tell the AI what you need in plain language. It uses the rinda-cli tool behind the scenes.

Workflow	Example
Buyer Search	"Find 50 cosmetics importers in the US with revenue over $5M"
Contact Enrichment	"Enrich the contacts for those buyers"
Email Campaigns	"Create a 6-step outreach sequence called 'US Cosmetics Q1'"
Reply Management	"Check my replies and tell me what to do"
Campaign Reports	"Show campaign performance for the last 30 days"

Workflows chain naturally — after each step, the AI suggests the next action and carries forward all relevant data (lead IDs, sequence IDs, etc.).

---

How It Works

Buyer Search

Specify industry, country, buyer type, and minimum revenue. The AI searches, scores each lead using weighted criteria (revenue, employee count, import history, product relevance), and presents a ranked table.

Contact Enrichment

Collects emails, phone numbers, LinkedIn profiles, and product info. Classifies leads by priority:

• High: Decision-maker email + strong product match
• Medium: Email found, unclear job title
• Low: Generic email only (info@, contact@)

Email Campaigns

Creates multi-step email sequences with AI-generated subject lines and bodies. Follows built-in guidelines:

• Subject lines under 60 characters, includes company name
• Body: Hook → Value → CTA
• Follow-ups add new value, progressively shorter

Reply Management

Classifies replies by intent and urgency:

Intent	Action
Meeting request	Respond immediately
Positive interest	Respond today
Question	Answer + follow up in 3 days
Not now	Reactivate in 90 days
Rejected	Update tags, remove from sequence
Unsubscribe	Process immediately

Campaign Reports

Funnel analysis, email metrics, hot leads list, and AI-generated insights. Compares against B2B export benchmarks (open rate >35% good, reply rate >10% good).

---

Architecture

This repo is a plugin marketplace — it can host multiple plugins under plugins/.

.claude-plugin/
  marketplace.json          # Central plugin registry
  plugin.json               # Marketplace metadata
plugins/
  rinda-ai/                 # B2B export sales plugin
    .claude-plugin/plugin.json
    hooks/hooks.json        # Auto-installs CLI + MCP server on session start
    skills/
      rinda/
        SKILL.md            # Main skill (CLI commands + workflows)
        references/
          buyer-qualification.md
          email-writing.md
          export-sales.md
bin/
  install.sh                # CLI + MCP server installer (cross-platform)
crates/
  cli/                      # rinda-cli (Rust, handles auth + API)
  sdk/                      # Auto-generated from OpenAPI spec
  mcp-server/               # rinda-mcp (Rust, HTTP MCP server with OAuth)

CLI

The rinda-cli binary handles authentication and all API calls. It installs to ~/.rinda/bin/rinda-cli automatically via a SessionStart hook.

MCP Server

The rinda-mcp binary provides an MCP tool interface auto-discovered by Claude Code via plugin.json. It installs to ~/.rinda/bin/rinda-mcp alongside the CLI. Claude Code launches it automatically on session start — no manual configuration needed inside the plugin.

See MCP Server (Standalone) below for usage outside the Claude Code plugin.

Command	Description
rinda-cli auth ensure-valid	Refresh token if expired
rinda-cli buyer search	Search for buyers
rinda-cli buyer enrich	Enrich a lead
rinda-cli reply check	Check email replies
rinda-cli campaign stats	Get campaign stats
rinda-cli sequence create	Create email sequence
rinda-cli email send	Send an email
rinda-cli config show	Show current configuration
rinda-cli order list	List order history

---

MCP Server (Remote)

The RINDA MCP server is hosted remotely and can be used with any MCP-compatible client that supports remote servers with OAuth (Claude Desktop, Cursor, etc.). No local installation required.

Server URLs:

Environment	URL
Beta (dev/test)	https://beta-mcp.rinda.ai
Production	https://mcp.rinda.ai

Configure your MCP client:

For Claude Desktop (macOS: ~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "rinda": {
      "url": "https://beta-mcp.rinda.ai"
    }
  }
}

The server supports OAuth 2.0 with dynamic client registration. On first connection, your MCP client will open a browser for Google authentication via RINDA. Tokens are managed automatically by the client.

OAuth discovery endpoints:

Endpoint	Description
/.well-known/oauth-authorization-server	RFC 8414 authorization server metadata
/.well-known/oauth-protected-resource	RFC 9728 protected resource metadata
/oauth/register	Dynamic client registration (RFC 7591)

---

FAQ

Do I need programming knowledge? No. Just describe what you want in plain language.

What industries/countries does it support? Any B2B export industry (cosmetics, food, machinery, electronics, etc.) and most countries worldwide.

Is my data secure? Auth tokens are stored locally at ~/.rinda/credentials.json (permissions 600). All communication uses HTTPS. Tokens auto-refresh and are deleted on logout.

What if my login expires? Tokens refresh automatically. Only re-authenticate if unused for 14+ days.

---

Technical Details

Item	Details
Plugin type	Claude Code Marketplace
Authentication	Google OAuth 2.0 via RINDA
CLI	Rust (cross-platform: Linux, macOS, Windows)
MCP Server	rinda-mcp (HTTP transport with OAuth, hosted remotely)
SDK	Auto-generated from OpenAPI spec via progenitor
Token validity	Access: 1 hour (auto-refresh), Refresh: 14 days
Credentials	CLI: ~/.rinda/credentials.json, MCP: OAuth session tokens
License	MIT

---

Docker / Deployment

Local Development

# 1. Copy and edit environment variables
cp .env.example .env

# 2. Build and start the MCP server
docker compose up

The server will be available at http://localhost:3000. Verify with:

curl http://localhost:3000/health
# {"status":"ok"}

Build the Docker Image

docker build -t rinda-mcp .
docker run -p 3000:3000 \
  -e RINDA_API_BASE_URL=https://alpha.rinda.ai \
  -e MCP_SERVER_URL=http://localhost:3000 \
  rinda-mcp

Dokploy Deployment

1. Create Application — In your Dokploy dashboard, create a new Application and point it at this repository.

2. Set build type to Dockerfile — Dokploy will use the Dockerfile at the repository root.

3. Configure environment variables — Set the following in the Dokploy environment panel:

   Variable	Example value	Description
   PORT	3000	Port the server listens on
   RUST_LOG	info	Log level (error, warn, info, debug)
   RINDA_API_BASE_URL	https://app.rinda.ai	RINDA API backend (use app for production)
   MCP_SERVER_URL	https://mcp.rinda.ai	Public URL of this server (used in OAuth callbacks)

4. Configure health check — Set the health check endpoint to GET /health.

5. Configure domain — Assign a domain (e.g. mcp.rinda.ai) and enable HTTPS. Update MCP_SERVER_URL to match.

6. Deploy — Trigger a deploy. Dokploy will build the multi-stage Docker image, run the container, and confirm health before routing traffic.

---

Support

• Website: rinda.ai
• Email: support@grinda.ai
• GitHub: FINGU-GRINDA/claude-rinda-plugin