For Developers

Your AI agent's cold email pipeline needs a deliverability safety net.

Q: How can an AI sales agent check inbox health before sending emails?

SimplyWarmup exposes a native MCP server and REST API. An AI agent can call the get_inbox_health tool before each outbound campaign cycle to check the current health score, reputation status, and pacing stage. If health drops below a configurable threshold — typically 75 — the agent should pause the campaign automatically and wait for health to recover above the safe threshold before resuming.

Q: What is the risk of an AI agent sending cold emails from a domain without warmup?

New or unwarmed domains are treated with high suspicion by Google and Microsoft. Even with correct SPF, DKIM, and DMARC setup, an inbox with no sending history will see 30–50% of emails routed to spam. For autonomous AI agents that send high volumes, this can trigger hard blocks within days — permanently damaging the domain's reputation and requiring 4–8 weeks of clean sending to recover.

Q: Does the MCP integration require changes to the AI agent's code?

Minimal changes. You add SimplyWarmup to your MCP server configuration file (typically mcp.json) with your API key. The MCP runtime then makes the get_inbox_health, pause_inbox, and resume_inbox tools available to your agent automatically. Adding a health check before each send batch requires approximately 5 lines of agent logic.

Autonomous AI SDRs can send thousands of emails per day. Without a reputation safeguard, one bad send cycle can permanently blacklist every domain in your outbound stack. SimplyWarmup is the circuit breaker your agent stack needs.

Connect an Inbox — $19/mo View API Docs

ai-sdr-runtime.log

// AI SDR cycle — [email protected]

Agent: Preparing 450-email campaign...

MCP call: get_inbox_health("[email protected]")

Response: score=94, status=HEALTHY

✓ Health confirmed. Proceeding.

Sending batch 1/9 (50 emails)...

✓ Delivered: 50 / 50

--- 6 hours later ---

MCP call: get_inbox_health("[email protected]")

Response: score=87, status=DEGRADED

⚠ Score below threshold. Halting.

MCP call: pause_inbox("[email protected]")

✓ Inbox paused. Warmup scheduled.

The problem with autonomous outbound at scale.

AI SDR agents are now capable of executing full outbound workflows — prospect research, email copy, follow-ups, and replies — with minimal human oversight. But they have no native understanding of email reputation, sender scores, or blacklist risks.

Without a deliverability layer, an AI agent will continue sending from a degraded inbox, accelerating the damage with every batch. By the time a human notices, the domain may be permanently blacklisted.

SimplyWarmup solves this by giving your AI agent a first-class deliverability tool via MCP — so health checks, campaign halts, and warmup triggers are native operations in your agent's decision loop.

What the MCP integration enables

Pre-send health checks

Agent calls get_inbox_health before every campaign batch. If score < 90, the send is automatically skipped.
Automatic pause on risk detection

Agent calls pause_inbox when health degrades, triggering a warmup recovery cycle before next campaign.
Fleet-level health monitoring

Agent calls list_inboxes to get health across all inboxes, selecting only healthy ones for a given send batch.
Programmatic provisioning

When a new SDR inbox is provisioned via your HR or CRM system, your agent calls the API to enroll it in warmup automatically — no human steps.

Getting SimplyWarmup into your agent stack.

Three steps from zero to production-ready deliverability infrastructure.

Connect your sending inboxes via OAuth

Sign up and connect your Google Workspace or Microsoft 365 sending inboxes via OAuth. Each inbox is validated against our acceptance criteria and enrolled in the Clean Pool. Warmup begins automatically on the same day.

Add SimplyWarmup to your MCP config

Add the SimplyWarmup MCP server to your agent's config file. Your API key is available in the developer workspace under /app/developer.

"simplywarmup": { "command": "npx", "args": ["-y", "@simplywarmup/mcp-server"], "env": { "SIMPLYWARMUP_API_KEY": "sw_live_..." } }

Add health checks to your campaign logic

Update your agent's send function to call get_inbox_health before each batch. If the score is below your threshold, call pause_inbox and skip the send. SimplyWarmup will recover the inbox automatically in the background.

FAQ

Questions from developers building autonomous outbound

How does an AI agent check inbox health before sending?

Connect SimplyWarmup to your agent's MCP runtime by adding it to your mcp.json config. The get_inbox_health tool returns a health score (0–100), reputation status, and pacing stage. If health drops below your threshold — typically 75 — implement pause logic before the send batch. This adds approximately 5 lines of agent code and removes human monitoring from your outbound loop entirely.

What's the risk of running autonomous cold email without inbox health monitoring?

Without health monitoring, an agent that continues sending from an at-risk inbox can trigger domain blacklisting within days. New or unwarmed domains see 30–50% of emails routed to spam even with correct authentication. Once a domain is blacklisted, recovery requires 4–8 weeks of clean sending — effectively halting all outbound for that period. Health checks are the circuit-breaker that prevents this.

Does the MCP integration require changes to existing agent code?

Minimal. Adding SimplyWarmup to your MCP configuration file takes under a minute. Your agent then has access to all health and warmup tools automatically. Adding a health check gate before campaign sends requires 5 lines of agent logic. The full TypeScript integration example is shown on the MCP Integration page.