MCP Server — Beta

Run outreach from
where you already work.
Or build your own agent.

RocketSDR's MCP server works two ways: plug it into Claude or ChatGPT to manage your pipeline in plain English, or use it as the outreach execution layer inside your own custom AI agent.

Built on the open Model Context Protocol standard
22
MCP tools available out of the box
2 min
Setup time from zero to first command
Custom agents you can build on top
For sales teams

Control RocketSDR from Claude or ChatGPT

Connect the MCP server to your AI assistant of choice. Ask questions, launch campaigns, triage your inbox — in plain English. No dashboard, no training, no context switching.

Claude Desktop & Claude.ai
ChatGPT with Plugins
2-minute setup via API key
For builders

Use RocketSDR as the execution layer in your own AI agent

Building a custom AI SDR, a RevOps automation, or an internal sales tool? Point any MCP-compatible agent framework at the RocketSDR server and get full outreach execution without building the infrastructure yourself.

Works with any MCP-compatible agent framework
Full read + write access to all 22 tools
No rate limits on tool calls — build at scale

For sales teams

Your full pipeline, in conversation

Six things you can do right now, from Claude or ChatGPT, without touching the dashboard.

🚀

Launch campaigns

Create a new outreach campaign, set your targeting, and start it — all from a single message to Claude or ChatGPT. No UI navigation required.

📬

Read your inbox

Ask your AI assistant to pull the latest replies, summarise sentiment, or surface any positive responses that need follow-up. Your inbox, accessible in conversation.

📊

Check campaign performance

Get send counts, open rates, reply rates, and A/B test results in natural language. Ask 'how is my Q2 outbound campaign doing?' and get a real answer.

👤

Manage personas & products

Create, list, or update the personas and product descriptions your campaigns use for personalisation — without ever opening the RocketSDR dashboard.

🔎

Preview leads

Before launching, ask your AI to preview what prospects would match a given targeting configuration. Validate your audience in conversation.

⏸️

Pause, resume & update

Pause a running campaign, update its status, skip a lead, or resume a contact from conversation. Full campaign control, no tab switching.

In practice

Talk to your pipeline like it's a colleague.

The MCP connector gives Claude and ChatGPT direct access to your RocketSDR workspace. No workarounds, no copy-pasting. You ask, the AI acts.

Works with Claude Desktop, Claude.ai, and ChatGPT

All 22 RocketSDR actions available as MCP tools

No dashboard required for day-to-day campaign management

Full read and write access — view, create, update, and control

Claude — RocketSDR MCP
U

How's the Q2 outbound campaign performing?

Your Q2 outbound campaign has sent 847 emails with a 68% open rate and 12.4% reply rate — 3× your baseline. 6 positive replies are waiting in your inbox, including one from a VP of Sales at a Series B fintech.

U

Start a new campaign targeting CTOs at climate tech companies raising Series A.

Done. I've created the campaign with people targeting scoped to CTO-level contacts at climate tech companies that raised Series A in the last 12 months. ~340 matching prospects. Want me to preview a sample before enabling?

For builders

RocketSDR as the outreach brain inside your agent.

If you're building a custom AI SDR, a RevOps automation, or an internal sales tool — RocketSDR's MCP server gives your agent full outreach execution capability. You write the logic. RocketSDR handles prospecting, research, sending, and reply tracking.

Internal AI SDR agents

Build an agent that monitors your CRM for new inbound leads, qualifies them against your ICP, and automatically triggers a personalised outreach campaign — zero human steps.

Slack-native campaign control

Wire up a Slack bot that lets your whole team launch campaigns, check reply rates, or skip a lead — directly from a channel. No one needs access to the RocketSDR dashboard.

Signal-to-send automation

Connect your own signal sources — product usage events, CRM stage changes, Clearbit webhooks — to an agent that decides when to start outreach and calls RocketSDR to execute.

Custom SDR workflows

Build an agent that handles your entire outbound loop: prospect identification, campaign creation, reply triage, follow-up scheduling. RocketSDR handles the execution; your agent handles the logic.

agent.py — using RocketSDR MCP tools
# Your agent decides when to trigger outreach.
# RocketSDR handles everything that happens next.
leads = mcp.call("preview_leads", targeting=icp_config)
if leads.count > 50:
mcp.call("create_campaign", persona="cto_v2", ...targeting)
# Check for positive replies and route to Slack
inbox = mcp.call("get_inbox")
positives = [m for m in inbox if m.sentiment == "positive"]
for reply in positives:
slack.notify(channel="#sales", message=reply)

Setup

Three steps to full access

01

Install the RocketSDR MCP connector

Add the RocketSDR MCP server to your Claude or ChatGPT configuration. It takes about two minutes and works with Claude Desktop, Claude.ai, and ChatGPT with Plugins enabled.

02

Authenticate with your RocketSDR account

Connect your RocketSDR workspace via API key. Your campaigns, inbox, and data are immediately accessible from your AI assistant.

03

Talk to your outreach pipeline

Ask in plain English. 'Show me my best-performing campaign this month.' 'Create a new campaign targeting VPs of Sales at Series B fintech companies.' 'What replies came in today?' The AI handles the rest.

MCP tools

22 tools. Full coverage.

Every action in the RocketSDR platform is exposed as a discrete MCP tool — available to your AI assistant or your custom agent with the same API.

list_campaigns
get_campaign
create_campaign
create_campaign_draft
update_campaign_status
get_campaign_leads
get_lead
skip_lead
resume_lead
get_inbox
get_lead_threads
get_thread_messages
list_personas
create_persona
list_products
create_product
list_senders
search_targeting
preview_leads
preview_email
get_campaign_ab_metrics
get_csv_upload_url
"

We were blasting the same pitch to every manufacturer on the list. RocketSDR started triggering campaigns when companies posted Quality Control Engineer roles — that's a company actively scaling manual inspection. That's our exact buyer. Reply rates doubled in the first week.

Aditya
Aditya
Founder, Frinks.ai — AI Quality Control

Use it. Build on it.
Either way, it's ready.

The RocketSDR MCP server is live in beta. Book a demo to see it connected, or talk to us about building on top of it.