Channel-safety for cold outreach: volume limits, SPF/DKIM/DMARC, spam-trigger auditing.
The channel-safety layer for cold outreach. Volume discipline, sender setup, spam auditing — so good messages actually arrive.
Disclaimer. Platform limits and filtering behavior change frequently; values are conservative practitioner ranges as of last verification, not platform guarantees. Legal notes (CAN-SPAM / GDPR / PECR / CASL) are orientation, not legal advice.
Most outbound fails before the copy is ever read. Burned domains, throttled LinkedIn accounts, and spam-foldered sends kill campaigns that any message audit would have passed. AI agents make this worse: they can generate 500 perfect emails an hour, and sending them is exactly how you lose the domain.
This MCP is the layer that keeps the channel alive. It does NOT send anything — it returns rules, checklists, and audits that an agent consults BEFORE and DURING a campaign.
The core diagnostic it encodes: deliverability problems masquerade as copy problems. A 1% reply rate usually isn't a bad hook — it's a foldered send. Fix order: infrastructure → list → message.
| Tool | What it returns |
|---|---|
get_channel_rules | Safe daily volumes, account setup, warmup protocol, kill signals, personalization floor, and legal basics per channel: email, LinkedIn, Instagram DM, X DM |
get_sender_setup_checklist | The 10-step cold-email infrastructure checklist: lookalike domains, SPF, DKIM, DMARC, mailboxes, warmup, list verification, tracking, compliance, monitoring |
audit_spam_triggers | Copy audit: 30+ flagged phrases, all-caps, exclamations, link count, attachments, money symbols, fake Re:/Fwd: — with a 0-100 risk score |
interpret_benchmark | Healthy/warning ranges for open rate, reply rate, bounce rate, LinkedIn acceptance, DM reply — with an optional direct read of your observed value |
diagnose_campaign | Given your observed metrics, names the failure layer (infrastructure vs list vs message) and the fix order |
get_full_pack | The complete library as one payload for full agent context |
// Campaign feels dead. Copy problem or channel problem?
mcp.call("diagnose_campaign", {
channel: "email", open_rate: 31, reply_rate: 0.8, bounce_rate: 4.2
});
// Returns: failure layer = LIST. Bounce 4.2% > 3% is actively damaging the
// domain; open 31% says sends are being foldered. Fix infrastructure and
// list BEFORE touching the copy.
mcp.call("audit_spam_triggers", { text: "URGENT: Don't miss this LIMITED TIME offer!!! Click here..." });
// Returns: HIGH RISK, phrase + structural hits itemized.
Pair with outbound-engine-mcp: that one makes the message worth sending; this one makes sure it lands.
Elisabeth Hitz — 10+ years of B2B enterprise sales experience across ad-tech, SaaS, media, and global hiring. Now building MCP servers for the AI agent ecosystem.
License: MIT
Source-derived launch command. Check the maintainer’s required arguments and credentials before running:
npx -y outreach-deliverability-mcpMerge this template into ~/Library/Application Support/Claude/claude_desktop_config.json. Keep existing servers. Add any arguments, credentials, and permissions required by the maintainer; this template has not been install-tested.
{
"mcpServers": {
"io-github-closermethod-outreach-deliverability-mcp": {
"command": "npx",
"args": [
"-y",
"outreach-deliverability-mcp"
]
}
}
}Restart Claude Desktop completely for changes to take effect. Confirm the server appears connected in the client’s tool list, then try a read-only example from its documentation.
Claude Desktop setup referenceoutreach-deliverability-mcpnpmio.github.closermethod/outreach-deliverability-mcp works with any MCP-compatible client. Copy the config snippet from the Configuration section above and add it to the file shown for your client, then restart the application.
~/Library/Application Support/Claude/claude_desktop_config.jsonRestart Claude Desktop completely for changes to take effect.~/.cursor/mcp.jsonRestart Cursor for changes to take effect..vscode/mcp.jsonReload VS Code window for changes to take effect.~/.codeium/windsurf/mcp_config.jsonRestart Windsurf for changes to take effect..mcp.jsonSave at the project root, then start Claude Code in that project and review the MCP server approval prompt. Keep real credentials out of shared files.