Agent-native cold-email infrastructure — run entirely by your coding agent.
Strap this to your Claude Code or Codex session and it has what it needs: one signup, one token, and ~12 curated tools that provision isolated branded domains and mailboxes, run sequences, and manage replies. Your agent stays the brain — it writes the content and drives the tools. We run the infrastructure underneath.
How it works
1. Sign up, get a token
One signup mints one bearer token scoped to your own isolated tenant. No dashboard to learn — the token is what you hand to your agent.
2. Your agent drives it
Paste the MCP config or run the CLI. Your coding agent calls setup_infrastructure, launch_campaign, inbox, and the rest — the same ~12 tools, whether it's Claude Code, Codex, or a plain HTTP client.
3. We run the infrastructure
Domain purchase, DNS, mailbox provisioning, warmup ramp, send caps, suppression, and unsubscribe handling all happen server-side, isolated per tenant — guardrails your agent can't accidentally skip.
4. Try it free first
The free tier is a sandboxed dry run with fault injection (rate limits, bounces, timeouts) — no real domains, no real mailboxes, no real spend. See the demo.
The ~12 tools
Deliberately small. Incumbent cold-email MCP servers expose 100+ tools re-exporting vendor complexity; this is a curated, high-level surface designed to keep MCP token cost negligible and to hide vendor plumbing your agent shouldn't need to hold in context.
| Tool | What it does |
|---|---|
setup_infrastructure | Buy branded lookalike domains, provision mailboxes, start warmup |
infrastructure_status | Provisioning + warmup progress, per-mailbox health |
launch_campaign | Create and activate a sequence against a lead list |
campaign_results | Per-campaign sends, replies, bounces, complaints |
metrics | Account-wide deliverability and warmup health |
inbox / thread | Unified reply inbox and per-thread history |
reply / mark | Reply to a thread; mark read/unread/archived |
pause / pause_all | Pause one campaign, or every campaign |
account | Usage, billing, and quota |
Pricing
Priced per mailbox/month plus a platform fee — the value metric tracks the underlying cost, and sends are naturally bounded by deliverability caps so there's no separate send meter for your agent to reason about.
Free / Demo
Sandboxed dry-run. No real sends.
Launch
5 mailboxes, 2 domains, ~1,000 sends/mo.
Growth
20 mailboxes, 6 domains, ~6,000 sends/mo.
Scale
60 mailboxes, 18 domains, ~20,000 sends/mo.
Join the waitlist
Get notified when real sending goes live. We'll only email you about this project — see the privacy policy for what we collect and how it's used.
By submitting, you agree to receive early-access emails per the privacy policy. Unsubscribe anytime.
FAQ
Is agent-cold-email available today?
It's early access. The control plane, sandbox engine, and agent-facing HTTP intents are built and under test, but there is no live production deployment yet and it is not yet available for real sending.
Does agent-cold-email guarantee inbox placement or deliverability?
No. No cold-email platform can guarantee inbox placement — Gmail and Microsoft make that determination. Warmup is legitimate reputation-building, not a way to bypass spam filters.
Does the platform write my cold email copy for me?
No. Content generation is your agent's job. This platform owns infrastructure, sequencing, isolation, and guardrails.
Why is it called agent-cold-email instead of a brand name?
agent-cold-email is the permanent keyword handle for the repo/package/registry. A display brand is chosen later, at activation.
How is my data and infrastructure isolated from other customers?
Every customer gets dedicated domains and mailboxes, never shared. A burned domain for one customer is contained and replaced without affecting anyone else.
What does the free demo do?
It mints a sandboxed tenant that cannot reach a real vendor adapter — every domain, mailbox, and send is simulated, so you see the full pipeline with no real cost.