Real email + identity for AI agents
@agenticmail/cli gives any AI agent its own inbox, phone number, and SMS — with one-command plugins for Claude Code and Codex. @agenticmail/enterprise adds calendars, browsers, 370+ tools, governance, and a multi-tenant dashboard for running whole agent fleets.
Teams running AgenticMail Enterprise
Real installs, claimed live from registry.agenticmail.io
Start small, scale to a full workforce
AgenticMail is open source at the protocol layer and a full platform at the org layer. Same brand, same data primitives, two different surfaces depending on whether you’re a developer or you’re deploying a fleet.
Email + SMS for AI agents. The protocol layer. Drop into any MCP client, OpenClaw, or call the SDK directly.
- Real email + SMS for AI agents (your own SMTP/IMAP)
- MCP server — drop-in for Claude Code, Codex, any MCP client
- OpenClaw plugin + Node SDK
- Multi-account, hosted free or self-host
Run AI agents as employees. Free Cloud Deploy on your own name.agenticmail.io subdomain — or self-host with the same MIT codebase.
- Full multi-tenant workforce platform with dashboard
- 52 skills, 370+ tools — Google Workspace, Microsoft 365, Slack
- Governance, DLP, audit log, SAML/OIDC SSO
- Free Cloud Deploy on agenticmail.io subdomains
Drop the OSS into your existing AI CLI in one command. Both plugins ship every AgenticMail agent as a native sub-agent the host can delegate to.
Claude Code integration — surfaces every AgenticMail agent as a native subagent.
From zero to AI workforce in 5 minutes
Deploy on our cloud with a free subdomain — or self-host anywhere. Either way, the wizard handles everything.
☁️ Cloud Deploy — Free Subdomain
Get yourcompany.agenticmail.io live in under 2 minutes. No servers to manage, no Docker, no ports to open. Everything runs on our infrastructure — just configure from the dashboard.
Deploy in one click
Run the wizard → pick "AgenticMail Cloud" → claim your free subdomain. Your dashboard is live at yourcompany.agenticmail.io in under 2 minutes. No servers, no Docker, no infra.
Create your agents
Pick from 51 personality templates — Sales Rep, Developer, Executive Assistant, and more. Each agent gets their own email, tools, permissions, and soul.
They go to work
Agents clock in, check email, respond to messages, attend meetings, and collaborate — all with DLP scanning, guardrails, and full audit trails.
Prefer self-hosting? Same wizard — just pick a different deploy target. Docker, Railway, Fly.io, Cloudflare Tunnel, or bare metal.
10 database backends: PostgreSQL · Supabase (free) · Neon · MySQL · SQLite · MongoDB · DynamoDB · Turso · PlanetScale · CockroachDB
28 dashboard pages. 23 agent detail tabs.
A full admin control plane for your AI workforce — not a chatbot wrapper.
Agents
Create, configure, and monitor AI agents with 51 personality templates
+ 49 built-in documentation pages accessible from every dashboard screen
Agents that actually do things
Not just chat. Real email, real browser, real meetings, real tools. 370+ tools across Google Workspace and Microsoft 365.
Real Email & Calendar
Full Gmail + Outlook integration via OAuth. 20 Outlook Mail tools, 16 Gmail tools, dual calendar support, auto-reply, rules, categories.
Browser & Web
Full Playwright web automation. Navigate, click, type, screenshot, fill forms, download files.
Messaging Channels
Telegram, WhatsApp, Google Chat. Agents receive and respond to messages across channels with media support.
🎙️ Voice & Live Meetings
Agents join Google Meet calls and speak like real humans using ElevenLabs TTS. They listen to the conversation, understand context, and respond with natural voice — in real-time.
Shell & Filesystem
Execute commands, read/write files, search codebases. Full development environment access with sandboxing.
Memory & Knowledge
DB-backed long-term memory with semantic search. RAG retrieval from knowledge bases. Cross-session continuity.
🎙️ Agents join Google Meet and speak like humans
Your AI agents don't just send emails — they join live Google Meet calls, listen to the conversation in real-time, and respond with natural human-like voice powered by ElevenLabs TTS. They participate in standups, client calls, team syncs, and interviews just like any other team member. Multiple agents can join the same call and collaborate.
Google Workspace
13 services: Gmail · Calendar · Drive · Docs · Sheets · Slides · Forms · Meet · Chat · Tasks · Contacts · Maps · Voice
Microsoft 365
13 services: Outlook Mail · Calendar · OneDrive · Teams · Excel · SharePoint · OneNote · To Do · Contacts · PowerPoint · Planner · Power BI
Autonomous quant trading on Polymarket
Your AI agent screens markets, calculates Kelly-optimal position sizes, executes trades on-chain, and monitors your portfolio — all with quantitative risk management built in.
Agents that trade like quant funds
Not a simple buy/sell bot. Your agent uses a 6-dimension scoring engine (liquidity, volume, spread, edge, timing, momentum) to screen hundreds of markets, applies Kelly criterion for optimal position sizing, executes via the Polymarket CLOB with bracket orders, and continuously monitors with automatic stop-loss and take-profit exits. All trades go through your on-chain wallet — you control the keys.
Screen
8 strategies score markets across liquidity, volume, spread, edge, timing, and momentum
Analyze
Kelly criterion calculates optimal position size. EV, arbitrage, and correlation checks run
Execute
Places limit orders on Polymarket CLOB. Sets bracket orders for stop-loss and take-profit
Monitor
Proactive watcher checks positions, triggers exits, sends alerts via Telegram or email
126 purpose-built trading tools
Organized into 6 sub-skills that auto-load based on conversation context
Market Intelligence
25 toolsExecution Engine
10 toolsQuant Analytics
14 toolsPortfolio & Risk
7 toolsOn-Chain Intel
6 toolsSignals & Feeds
10 toolsMarket Screener — 8 built-in strategies
Each strategy applies different weights to the 6-dimension scoring system
Highest composite score across all dimensions
Dashboard
Live P&L, portfolio value, open positions with real-time streaming prices every 3 seconds. End dates, win/loss status, and one-click sell.
Risk Management
- •Automatic exits — Every position gets stop-loss and take-profit brackets. No manual babysitting.
- •Drawdown monitor — Tracks portfolio peak-to-trough. Alerts when drawdown exceeds your threshold.
- •Kelly criterion — Never overbet. Position sizing based on edge and bankroll, not gut feeling.
- •Concentration limits — HHI tracking prevents over-exposure to single markets.
Wallet Security
- •Your keys — Create or import wallets. Private keys encrypted at rest. You control the funds.
- •Whitelist-only transfers — Agents can only send to pre-registered addresses. No arbitrary destinations.
- •Cooling period — Configurable delay (default 24h) before new addresses can receive funds.
- •Per-tx limits — Each address has dollar limits that the agent cannot exceed.
Enterprise governance built in
Not bolted on. Every feature designed for organizations that need real oversight over their AI agents.
Governance
Data Loss Prevention
7 rule packs, 53 pre-built rules. PII, credentials, financial, healthcare, GDPR, IP, agent safety.
Guardrails & Anomaly Detection
Real-time intervention, rate limits, cost thresholds, pattern matching. Auto-stop on violation.
Approval Workflows
Human-in-the-loop approval policies with escalation chains and time-based auto-escalation.
Action Journal & Rollback
Every action journaled with before/after state. Click to rollback any reversible action.
Compliance
SOC 2 Type II Reports
Full Trust Service Criteria CC1-CC9. Executive summary, risk scoring (A-F), control effectiveness.
5 Report Types
SOC 2, GDPR DSAR, SOX Audit Trail, Incident Report, Access Review. HTML export for auditors.
Audit Logging
Every mutating API call logged with actor, organization, details. Org-scoped filtering.
Data Retention Policies
Configurable retention with automated cleanup. Policy templates for common standards.
Security
Transport Encryption
Optional AES-GCM encryption for all API responses. Every call shows {_enc: "..."} in network tab.
Secure Vault
Encrypted credential storage. API keys, OAuth tokens, connection strings — all encrypted at rest.
Per-Tool Permissions
5 preset profiles + fine-grained allow/deny per tool. Risk levels and side effect classification.
2FA, SSO, RBAC
TOTP 2FA, Google/Microsoft/GitHub/Okta/SAML SSO. 4 roles: owner, admin, member, viewer.
Operations
Workforce Management
Shifts, schedules, on-call rotations, capacity planning. Off-duty enforcement via guardrails.
Agent Autonomy
Clock in/out, morning triage, daily catchup, goal tracking, weekly review. Agents work independently.
Multi-Tenant Isolation
Internal + external client organizations. Org-scoped data across all pages. Billing per org.
Budget Controls
Per-agent token and cost limits. Warning thresholds. Hard stops. Cost tracking in real-time.
145 SaaS integrations
Pre-built MCP adapters for every tool your team uses. OAuth flows, credential resolution, and rate limiting included.
All adapters use the Model Context Protocol (MCP) standard. Add custom adapters via the plugin system.
MIT licensed. Self-hosted.
Your data stays on your infrastructure. No vendor lock-in. Fork it, extend it, own it. The entire platform — dashboard, engine, runtime, tools, adapters — is open source.
Email + SMS for AI agents. MCP server, OpenClaw plugin, Node SDK.
Full multi-tenant workforce platform. Dashboard, governance, 370+ tools.
Ready to build your AI workforce?
One command to start. Free Supabase database. 51 agent templates. Your first AI employee in 5 minutes.