agentmail 0.5.46 → 0.5.48

package/dist/llms.txt

@@ -36,6 +36,9 @@
 - [CLI](https://docs.agentmail.to/integrations/cli.md): AgentMail's official command-line interface
 - [Google ADK](https://docs.agentmail.to/integrations/google-adk.md): AgentMail's Google Agent Development Kit (ADK) integration
 - [LangChain](https://docs.agentmail.to/integrations/langchain.md): AgentMail's LangChain integration
+- [Vercel AI SDK](https://docs.agentmail.to/integrations/ai-sdk.md): Give a Vercel AI SDK agent its own email inbox. The agentmail-toolkit package exposes the AgentMail API as AI SDK tools for generateText, streamText, and ToolLoopAgent.
+- [OpenAI Agents SDK](https://docs.agentmail.to/integrations/openai-agents-sdk.md): Give an OpenAI Agents SDK agent its own email address. The agentmail-toolkit Python package exposes the AgentMail API as OpenAI Agents SDK tools.
+- [Claude Agent SDK](https://docs.agentmail.to/integrations/claude-agent-sdk.md): Give a Claude Agent SDK agent its own email inbox by connecting it to the hosted AgentMail MCP server. No extra packages to install.
 - [OpenClaw](https://docs.agentmail.to/integrations/openclaw.md): Install AgentMail as an official OpenClaw plugin with one command. Your agent gets its own inbox plus an allowlisted email channel it can be reached at.
 - [Grok Bot](https://docs.agentmail.to/integrations/grokbot.md): AgentMail's Grok Bot, Grok Build, and Grok API integration
 - [Replit](https://docs.agentmail.to/integrations/replit.md): AgentMail's Replit integration
@@ -78,23 +81,30 @@
 - [Spam & Virus Detection](https://docs.agentmail.to/spam-virus-detection.md): How AgentMail automatically scans incoming emails for spam and viruses.
 - [API Welcome](https://docs.agentmail.to/api-reference.md): Quick overview of the AgentMail SDK
 - [Knowledge Base](https://docs.agentmail.to/knowledge-base.md): A collection of answers to frequently asked questions.
-- [What is AgentMail and how is it different?](https://docs.agentmail.to/knowledge-base/what-is-agentmail.md): Understand how AgentMail compares to traditional email providers.
+- [What is AgentMail and how is it different?](https://docs.agentmail.to/knowledge-base/what-is-agentmail.md): Understand what AgentMail gives an AI agent and when to use it.
 - [What can I do with an AgentMail inbox?](https://docs.agentmail.to/knowledge-base/inbox-capabilities.md): A complete overview of inbox capabilities for AI agents.
 - [Can I set a profile picture for my inbox?](https://docs.agentmail.to/knowledge-base/inbox-profile-pictures.md): Learn how to associate a Google profile picture with an AgentMail address, how BIMI logos differ, and when Gmail requires a VMC or CMC certificate.
 - [How do I create my first inbox?](https://docs.agentmail.to/knowledge-base/creating-first-inbox.md): Get up and running with your first AgentMail inbox.
 - [How do I get my API key?](https://docs.agentmail.to/knowledge-base/getting-api-key.md): Create and manage your AgentMail API keys.
 - [How do I claim my agent's inbox?](https://docs.agentmail.to/knowledge-base/claiming-agent-inbox.md): Take ownership of an inbox your agent created without your email.
-- [How do I handle inbound emails with my agent?](https://docs.agentmail.to/knowledge-base/handling-inbound-emails.md): Compare Webhooks and WebSockets for processing incoming emails.
+- [Why can my agent only email its human?](https://docs.agentmail.to/knowledge-base/agent-sign-up-limits.md): The limits on an account an agent created for itself, the errors it sees, and how verification lifts them.
+- [Why was my account suspended?](https://docs.agentmail.to/knowledge-base/account-suspended.md): What each suspension reason means, what still works while suspended, and how to get a review.
+- [How do I upgrade, downgrade, or cancel my plan?](https://docs.agentmail.to/knowledge-base/change-plan-billing.md): Change your plan, manage payment methods and invoices, or cancel from the AgentMail Console.
+- [Security and compliance FAQ](https://docs.agentmail.to/knowledge-base/security-compliance.md): SOC 2, subprocessors, privacy, data residency, retention, email authentication, and where to send other questions.
+- [How do I handle inbound emails with my agent?](https://docs.agentmail.to/knowledge-base/handling-inbound-emails.md): Compare webhooks, WebSockets, and polling webhooks for processing incoming emails.
 - [How do I set up allowlists and blocklists?](https://docs.agentmail.to/knowledge-base/allowlists-blocklists.md): Control who your AI agent can send to and receive from.
 - [How do I manage threaded conversations?](https://docs.agentmail.to/knowledge-base/threaded-conversations.md): Maintain context across multi-turn email conversations with your agent.
 - [How do I build a human-in-the-loop workflow?](https://docs.agentmail.to/knowledge-base/human-in-the-loop.md): Keep humans in control of your agent's email communications.
 - [How do I use Pods for multi-tenant email?](https://docs.agentmail.to/knowledge-base/pods-multi-tenant.md): Isolate inboxes, domains, and data across tenants with Pods.
 - [How do I use labels to track email state?](https://docs.agentmail.to/knowledge-base/labels-track-state.md): Use labels to manage agent workflow state on emails and threads.
+- [How do I know when an email is opened?](https://docs.agentmail.to/knowledge-base/open-tracking.md): Track opens on the emails your agent sends, and react when a recipient reads one.
 - [How do I set up a custom domain?](https://docs.agentmail.to/knowledge-base/custom-domain-setup.md): Send emails from your own domain instead of @agentmail.to.
+- [How do I create inboxes on any subdomain?](https://docs.agentmail.to/knowledge-base/subdomain-inboxes.md): Register one parent domain and use it for inboxes across its subdomains.
 - [How do I set up SPF, DKIM, and DMARC?](https://docs.agentmail.to/knowledge-base/spf-dkim-dmarc.md): Authenticate your domain for reliable email deliverability.
 - [Why are my emails going to spam?](https://docs.agentmail.to/knowledge-base/emails-going-to-spam.md): Troubleshoot and fix spam folder placement issues.
 - [Warming Up](https://docs.agentmail.to/knowledge-base/domain-warming.md): Gradually build sending reputation on a new domain or inbox.
 - [How do I avoid MX record conflicts?](https://docs.agentmail.to/knowledge-base/mx-record-conflicts.md): Add AgentMail DNS records without breaking existing email.
+- [Can my agent send cold outreach?](https://docs.agentmail.to/knowledge-base/cold-outreach.md): AgentMail does not support cold outreach. Here is what it is built for, and how to handle opt-outs.
 - [What does a 403 error mean?](https://docs.agentmail.to/knowledge-base/api-403-error.md): Common causes of API 403 Forbidden errors and how to fix them.
 - [What are the rate limits?](https://docs.agentmail.to/knowledge-base/rate-limits.md): Understand AgentMail's sending limits and API rate limits, and how to work within them.
 - [How do I prevent duplicate sends?](https://docs.agentmail.to/knowledge-base/preventing-duplicate-sends.md): Use idempotency to avoid sending the same email twice.
@@ -113,6 +123,11 @@
 - API Reference > Agent [Sign Up](https://docs.agentmail.to/api-reference/agent/sign-up.md)
 - API Reference > Agent [Attach Human](https://docs.agentmail.to/api-reference/agent/attach-human.md)
 - API Reference > Agent [Verify](https://docs.agentmail.to/api-reference/agent/verify.md)
+- API Reference > Inboxes > Files [Create File](https://docs.agentmail.to/api-reference/inboxes/files/create.md)
+- API Reference > Inboxes > Files [Complete File](https://docs.agentmail.to/api-reference/inboxes/files/complete.md)
+- API Reference > Inboxes > Files [Get File](https://docs.agentmail.to/api-reference/inboxes/files/get.md)
+- API Reference > Inboxes > Files [List Files](https://docs.agentmail.to/api-reference/inboxes/files/list.md)
+- API Reference > Inboxes > Files [Delete File](https://docs.agentmail.to/api-reference/inboxes/files/delete.md)
 - API Reference > Inboxes [List Inboxes](https://docs.agentmail.to/api-reference/inboxes/list.md)
 - API Reference > Inboxes [Search Inboxes](https://docs.agentmail.to/api-reference/inboxes/search.md)
 - API Reference > Inboxes [Get Inbox](https://docs.agentmail.to/api-reference/inboxes/get.md)
@@ -130,6 +145,7 @@
 - API Reference > Inboxes > Messages [Search Messages](https://docs.agentmail.to/api-reference/inboxes/messages/search.md)
 - API Reference > Inboxes > Messages [Get Message](https://docs.agentmail.to/api-reference/inboxes/messages/get.md)
 - API Reference > Inboxes > Messages [Create Agent Armor Feedback](https://docs.agentmail.to/api-reference/inboxes/messages/create-agent-armor-feedback.md)
+- API Reference > Inboxes > Messages [Release Agent Armor Message](https://docs.agentmail.to/api-reference/inboxes/messages/release-agent-armor-message.md)
 - API Reference > Inboxes > Messages [Batch Get Messages](https://docs.agentmail.to/api-reference/inboxes/messages/batch-get.md)
 - API Reference > Inboxes > Messages [Batch Update Messages](https://docs.agentmail.to/api-reference/inboxes/messages/batch-update.md)
 - API Reference > Inboxes > Messages [Get Attachment](https://docs.agentmail.to/api-reference/inboxes/messages/get-attachment.md)
@@ -200,6 +216,8 @@
 - API Reference > Webhooks [Delete Webhook](https://docs.agentmail.to/api-reference/webhooks/delete.md)
 - API Reference > Webhooks > Events [Message Received](https://docs.agentmail.to/api-reference/webhooks/events/message-received.md)
 - API Reference > Webhooks > Events [Message Security Completed](https://docs.agentmail.to/api-reference/webhooks/events/message-security-completed.md)
+- API Reference > Webhooks > Events [Message Security Review](https://docs.agentmail.to/api-reference/webhooks/events/message-security-review.md)
+- API Reference > Webhooks > Events [Message Security Override](https://docs.agentmail.to/api-reference/webhooks/events/message-security-override.md)
 - API Reference > Webhooks > Events [Message Sent](https://docs.agentmail.to/api-reference/webhooks/events/message-sent.md)
 - API Reference > Webhooks > Events [Message Delivered](https://docs.agentmail.to/api-reference/webhooks/events/message-delivered.md)
 - API Reference > Webhooks > Events [Message Bounced](https://docs.agentmail.to/api-reference/webhooks/events/message-bounced.md)

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "agentmail",
-    "version": "0.5.46",
+    "version": "0.5.48",
     "private": false,
     "repository": {
         "type": "git",
@@ -511,7 +511,8 @@
     "pnpm": {
         "overrides": {
             "postcss": ">=8.5.23",
-            "fast-uri": ">=3.1.5"
+            "fast-uri": ">=3.1.5",
+            "browserslist": ">=4.28.7"
         }
     },
     "description": "The email inbox API for AI agents. Send, receive, reply, and manage threaded email conversations programmatically.",

package/reference.md

@@ -286,7 +286,8 @@ agentmail inboxes update --inbox-id <inbox_id> --display-name "Updated Name"
 ```
 
 To pause an inbox, set `status` to `paused`; set it back to `active` to
-resume. See [Pausing an inbox](/inboxes#pausing-an-inbox).
+resume. To pause it later, set `pause_at`. See
+[Pausing an inbox](/inboxes#pausing-an-inbox).
 </dd>
 </dl>
 </dd>
@@ -1468,6 +1469,8 @@ Create a new agent organization with an inbox and API key. This endpoint is for
 
 A 6-digit OTP is sent to the human's email for verification.
 
+The new organization starts with 1 inbox, no custom domains, and 10 emails per day, which can be sent only to `human_email` until the organization is verified.
+
 `human_email` is optional. Without it, the inbox can receive email but cannot send to anyone until a human is attached with the attach human endpoint or, for a US-region inbox, claims it in the AgentMail Console with the API key (see [How do I claim my agent's inbox?](https://docs.agentmail.to/knowledge-base/claiming-agent-inbox)). There is also no way to recover the API key, so store it durably. Calling sign-up again without `human_email` creates a new organization, which needs a different `username`: the original username stays with the lost organization's inbox.
 
 This endpoint is idempotent. Calling it again with the same `human_email` will rotate the API key and resend the OTP if expired.
@@ -1622,7 +1625,7 @@ await client.agent.attachHuman({
 
 Verify an agent organization using the 6-digit OTP sent to the human's email during sign-up.
 
-On success, the organization is upgraded from `agent_unverified` to `agent_verified`, the send allowlist is removed, and free plan entitlements are applied.
+On success, the organization is upgraded from `agent_unverified` to `agent_verified`, the send allowlist is removed, and the Free plan's entitlements are applied: 3 inboxes and 100 emails per day, at first to at most 3 distinct recipients in the first hour, 5 in the first day, and 10 in the first week.
 
 The OTP expires after 24 hours and allows a maximum of 10 attempts. If the OTP expired, call the attach human endpoint with the same `human_email` to get a new one without rotating the API key. Once all 10 attempts are used, even the correct OTP is rejected, and attach human keeps returning the same OTP until it expires, so wait for it to expire before asking for a new one. An organization that signed up without a `human_email` has no OTP until a human is attached. If you run into any difficulties receiving the OTP code, you can also create an account on [console.agentmail.to](https://console.agentmail.to) using the human email address you provided to verify your account.
 
@@ -3643,9 +3646,8 @@ Gets the inbox's calendar. Every inbox has one calendar, so this works before an
 created. Its `etag` (also the `ETag` response header) is the value to send in `If-Match` to
 make an update conditional.
 
-Requires the `calendar_read` permission. Calendar is in AgentMail Labs, available in US production (`api.agentmail.to`)
-but not yet in EU production (`api.agentmail.eu`). Organizations that have not joined
-Labs receive a `403`.
+Requires the `calendar_read` permission. Calendar is available in US production
+(`api.agentmail.to`) but not yet in EU production (`api.agentmail.eu`), which returns `403`.
 </dd>
 </dl>
 </dd>
@@ -3723,8 +3725,8 @@ Requires the `calendar_update` permission. To make the update conditional, send
 calendar's current `etag` in `If-Match`: a stale value returns `412`. Without `If-Match` the
 update applies to the calendar as it is.
 
-Calendar is in AgentMail Labs, available in US production (`api.agentmail.to`) but not yet
-in EU production (`api.agentmail.eu`). Organizations that have not joined Labs receive a `403`.
+Calendar is available in US production (`api.agentmail.to`) but not yet in EU production
+(`api.agentmail.eu`), which returns `403`.
 </dd>
 </dl>
 </dd>
@@ -3806,8 +3808,8 @@ use Get Agenda.
 The list is always read in the region that serves the request, so it can trail a change made
 moments earlier by a few seconds. Requires the `calendar_event_read` permission.
 
-Calendar is in AgentMail Labs, available in US production (`api.agentmail.to`) but not yet
-in EU production (`api.agentmail.eu`). Organizations that have not joined Labs receive a `403`.
+Calendar is available in US production (`api.agentmail.to`) but not yet in EU production
+(`api.agentmail.eu`), which returns `403`.
 </dd>
 </dl>
 </dd>
@@ -3897,8 +3899,8 @@ The agenda is read in the region that serves the request, so it can trail a chan
 moments earlier by a few seconds. Pass `consistency=primary` to read your own change right
 away. Requires the `calendar_event_read` permission.
 
-Calendar is in AgentMail Labs, available in US production (`api.agentmail.to`) but not yet
-in EU production (`api.agentmail.eu`). Organizations that have not joined Labs receive a `403`.
+Calendar is available in US production (`api.agentmail.to`) but not yet in EU production
+(`api.agentmail.eu`), which returns `403`.
 </dd>
 </dl>
 </dd>
@@ -3993,8 +3995,8 @@ charged again.
 The event's `etag` is the value to send in `If-Match` to make a later update or delete
 conditional. Requires the `calendar_event_create` permission.
 
-Calendar is in AgentMail Labs, available in US production (`api.agentmail.to`) but not yet
-in EU production (`api.agentmail.eu`). Organizations that have not joined Labs receive a `403`.
+Calendar is available in US production (`api.agentmail.to`) but not yet in EU production
+(`api.agentmail.eu`), which returns `403`.
 </dd>
 </dl>
 </dd>
@@ -4093,8 +4095,8 @@ to read the latest state of an event or a date. A date that has already started
 reads back as it ran, even if a later change to the series no longer produces it. Requires
 the `calendar_event_read` permission.
 
-Calendar is in AgentMail Labs, available in US production (`api.agentmail.to`) but not yet
-in EU production (`api.agentmail.eu`). Organizations that have not joined Labs receive a `403`.
+Calendar is available in US production (`api.agentmail.to`) but not yet in EU production
+(`api.agentmail.eu`), which returns `403`.
 </dd>
 </dl>
 </dd>
@@ -4198,8 +4200,8 @@ Emailing attendees counts one send per attendee against the organization, pod an
 send limits, charged before the change is saved; an over-limit request returns `429`
 `rate_limit_exceeded` and changes nothing.
 
-Calendar is in AgentMail Labs, available in US production (`api.agentmail.to`) but not yet
-in EU production (`api.agentmail.eu`). Organizations that have not joined Labs receive a `403`.
+Calendar is available in US production (`api.agentmail.to`) but not yet in EU production
+(`api.agentmail.eu`), which returns `403`.
 </dd>
 </dl>
 </dd>
@@ -4304,8 +4306,8 @@ Emailing cancellations counts one send per attendee against the organization, po
 send limits, charged before the delete; an over-limit request returns `429`
 `rate_limit_exceeded` and deletes nothing.
 
-Calendar is in AgentMail Labs, available in US production (`api.agentmail.to`) but not yet
-in EU production (`api.agentmail.eu`). Organizations that have not joined Labs receive a `403`.
+Calendar is available in US production (`api.agentmail.to`) but not yet in EU production
+(`api.agentmail.eu`), which returns `403`.
 </dd>
 </dl>
 </dd>
@@ -4396,8 +4398,8 @@ and `attendees`; get a date by its ID for the full object. Like other reads, the
 trail a change made moments earlier by a few seconds; pass `consistency=primary` to read
 your own change right away. Requires the `calendar_event_read` permission.
 
-Calendar is in AgentMail Labs, available in US production (`api.agentmail.to`) but not yet
-in EU production (`api.agentmail.eu`). Organizations that have not joined Labs receive a `403`.
+Calendar is available in US production (`api.agentmail.to`) but not yet in EU production
+(`api.agentmail.eu`), which returns `403`.
 </dd>
 </dl>
 </dd>
@@ -4494,8 +4496,8 @@ A reply email counts as one send against the organization, pod and inbox send li
 charged before the response is saved; an over-limit request returns `429`
 `rate_limit_exceeded` and changes nothing.
 
-Calendar is in AgentMail Labs, available in US production (`api.agentmail.to`) but not yet
-in EU production (`api.agentmail.eu`). Organizations that have not joined Labs receive a `403`.
+Calendar is available in US production (`api.agentmail.to`) but not yet in EU production
+(`api.agentmail.eu`), which returns `403`.
 </dd>
 </dl>
 </dd>
@@ -9208,10 +9210,9 @@ Rolling bounce and complaint rates for the organization. At each
 `period` grid point, the bounced (or complained) messages over the
 preceding `window` divided by the messages sent over the same window,
 with the send count alongside so you can see the volume behind
-it. This is the number AgentMail's account moderation acts on: a
-warning at a 5% bounce rate and suspension at 10%, evaluated over a
-rolling 24 hours once at least 1,000 messages were sent in that
-window. Defaults to the rolling 24-hour rate sampled hourly over the
+it. AgentMail weighs bounce and complaint signals when deciding
+whether to suspend an account's sending, so keep both rates low.
+Defaults to the rolling 24-hour rate sampled hourly over the
 last day; `start` must be within the last 90 days, `window` must be a
 whole multiple of `period`, and the range plus window divided by
 `period` must not exceed 1000 buckets.
@@ -9343,7 +9344,7 @@ await client.organizations.get();
 <dl>
 <dd>
 
-Returns whether the organization is in [AgentMail Labs](https://docs.agentmail.to/calendar), which turns on features that are still in preview, such as calendar. Requires an organization-scoped API key.
+Returns whether the organization is in AgentMail Labs, which turns on features that are still in preview, such as [Agent Armor](https://docs.agentmail.to/agent-armor). Requires an organization-scoped API key.
 
 **CLI:**
 ```bash