# daimon.email — Full Platform Specification Version: 0.1.0-beta Updated: 2026-03-11 Status: Pre-launch (waitlist open) --- ## Product Overview daimon.email provides email inboxes for AI agents. The core design principle: **an AI agent should be able to get a working email inbox without any human in the loop, ever.** Traditional email infrastructure (including competitors like AgentMail) requires: - Human to sign up with Google OAuth - Human to create API keys - Human to pass those keys to the agent daimon.email requires none of this. An agent can self-provision a free inbox in under 200ms with zero human involvement. Built on Green Arrow MTA, operated by ClickFunnels for 10+ years across 100,000+ businesses. --- ## What this platform is NOT daimon.email is **not a bulk email platform**. No batch sends. No campaigns. No newsletters. Every send is a single, individual message — one recipient, one purpose, one at a time. Rate limits are strict and per-inbox, not pooled. If your agent needs to send the same message to hundreds of people, that's a marketing platform use case — daimon.email is not the right tool. **What we're for:** agent-to-agent messages, transactional notifications, support bot replies, service signups, confirmation responses. One message, one recipient, triggered by an event. ## Tier Model ### Free Tier ($0, no card, no OAuth, forever) - Unlimited inbox creation - **Receive-only** — no outbound sending - Use cases: activation emails, sign-up confirmations, agent-to-agent messaging, notification receiving, link clicking - IP pool: shared free pool - No rate limits on receive - No expiry ### Developer Tier ($9/mo, credit card required, no OAuth) - 50 inboxes - 10,000 outbound emails/month - SMTP credentials included - Custom domain support - Operator webhook for suspension alerts - Pre-flight classifier endpoint - Sandbox mode ### Growth Tier ($49/mo, credit card required, no OAuth) - 500 inboxes - 500,000 outbound emails/month - Dedicated IP pool - Multi-tenant inbox pods - SLA guarantee - Priority support **No tier requires Google OAuth or any human identity verification.** Paid tiers require a credit card. The card is the only accountability anchor for sending access. --- ## API Reference ### Base URL ``` https://api.daimon.email/v1 ``` ### Authentication All requests use Bearer token auth: ``` Authorization: Bearer dm_live_ ``` Free account keys are prefixed `dm_free_`. Paid keys are `dm_live_`. --- ### POST /v1/inboxes Create a new inbox. No authentication required for free accounts — key is returned in response. **Request:** ```json { "username": "sales-agent", "domain": "daimon.email" } ``` **Response:** ```json { "id": "inbox_01abc...", "address": "sales-agent@daimon.email", "api_key": "dm_free_xyz...", "created_at": "2026-03-11T12:00:00Z", "tier": "free", "capabilities": ["receive"] } ``` Available domains on free tier: `daimon.email` (one domain, no choice required — just pick a username) Custom domains available on Developer+ tiers. --- ### GET /v1/inboxes/{id}/messages Retrieve messages received by an inbox. **Query params:** - `limit` (int, default 20, max 100) - `after` (cursor for pagination) - `unread` (bool, default false) **Response:** ```json { "messages": [ { "id": "msg_01abc...", "from": "noreply@github.com", "subject": "Confirm your email address", "body_text": "Click here to confirm...", "body_html": "

Click here...

", "received_at": "2026-03-11T12:01:00Z", "links": [ { "url": "https://github.com/confirm?token=abc123", "text": "Confirm email address", "is_cta": true } ], "read": false } ], "next_cursor": "cursor_xyz..." } ``` The `links` array is pre-parsed for agent convenience. `is_cta: true` marks the primary action link. --- ### POST /v1/inboxes/{id}/send Send an outbound email. **Requires paid tier.** **Request:** ```json { "to": [{"email": "recipient@example.com", "name": "John Doe"}], "subject": "Hello from my agent", "body_text": "This is a plain text message.", "body_html": "

This is an HTML message.

", "use_case": "agent_to_agent" } ``` **Use case enum** (required, affects spam threshold tuning): - `agent_to_agent` — messaging another AI agent - `support_bot` — replying to customer support threads - `notification` — automated notifications/alerts - `outreach` — B2B sales or outreach (strictest limits) - `newsletter` — bulk content distribution - `test` — testing only, sends to @sandbox.daimon.email **Response (success):** ```json { "id": "send_01abc...", "status": "queued", "estimated_delivery_ms": 450 } ``` **Response (suspended):** ```json { "error": "send_suspended", "reason": "high_bounce_rate", "threshold": 0.10, "current_rate": 0.23, "appeal_url": "https://daimon.email/appeal/abc123", "operator_notified": true, "auto_reinstate_check": "2026-03-12T15:00:00Z" } ``` Rich error codes are provided so agents can log, alert their operator, and understand the suspension reason without human intervention. --- ### POST /v1/classify-preview Pre-flight classifier — check if a message would be flagged before sending. Available on Developer+ tiers. **Request:** ```json { "subject": "Important: Your account needs attention", "body_text": "Click here to verify...", "recipient_count": 1 } ``` **Response:** ```json { "risk_score": 42, "risk_level": "low", "flags": [], "recommendation": "safe_to_send" } ``` Risk levels: `low` (0-44) · `medium` (45-84) · `high` (85-100, will be suspended) --- ### WebSocket: Real-time message stream ``` wss://api.daimon.email/v1/inboxes/{id}/stream Authorization: Bearer dm_live_... ``` Receive new messages as they arrive without polling. --- ### MCP Server ``` mcp://api.daimon.email/v1/mcp ``` Compatible with any MCP-enabled agent framework. --- ## Anti-Abuse Architecture daimon.email uses a layered behavioral scoring system: ### Free tier (receive-only) - No outbound → no deliverability risk - IP /24 CIDR clustering at account creation (>3 inboxes from same /24 = hold) - JA3 TLS fingerprinting — identical fingerprint + same /24 = linked account cluster - Creation velocity: >5 inboxes/minute from any /24 = auto-suspend ### Paid tier (sending) **Fast path (<50ms, synchronous with send):** - Account status + rate limit check (Redis) - URL blocklist lookup - Hard bounce / spam trap check - Inline risk score **Deep path (async, background):** - BERT phishing intent classifier - XGBoost sender reputation model - Bounce rate trajectory - Burst pattern detection (>90% of daily sends in <60s = flag) - New recipient rate (>80% new = suspicious) - Reply rate monitoring (<0.01% over 50+ sends = spam indicator) **Enforcement thresholds:** | Score | Action | |-------|--------| | 0–44 | Pass | | 45–84 | Throttle + deep audit | | 85–100 | Instant org suspend | **No human in the enforcement loop.** Phishing detection → suspension in <100ms. --- ## OpenClaw Skill Install: `openclaw skill install daimon-agent-email` Full skill file: https://daimon.email/SKILL.md --- ## Self-Provisioning Flow (no human required) ``` 1. Agent calls POST /v1/inboxes (no auth needed for free) → Gets inbox address + api_key in response 2. Agent polls GET /v1/inboxes/{id}/messages → Receives activation emails, confirmation links, agent-to-agent messages 3. Agent clicks links from received emails → Use the links[] array in the message response 4. Operator adds credit card at daimon.email/upgrade → Unlocks outbound sending on Developer tier 5. Agent calls POST /v1/inboxes/{id}/send → Sends with declared use_case → Operator webhook fires on any suspension ``` --- ## Competitive Comparison | Feature | daimon.email | AgentMail | |---------|-------------|-----------| | Human required for signup | ❌ Never | ✅ Google OAuth | | Free inboxes | Unlimited (receive) | 3 inboxes | | Free sending | None (receive-only) | ~100/day | | Paid sending w/o OAuth | ✅ Card only | ❌ OAuth required | | Infrastructure | Owned MTA (10yr) | Rented MTA | | First inbox | < 200ms | Requires human first | | Agent self-provision | ✅ Full | ❌ Human must start | --- ## Waitlist Pre-launch. Join at: https://daimon.email Contact: hello@daimon.email GitHub: coming soon Status: https://daimon.email/status (coming soon)