# Runearc (remote · mcp.runearc.com)

Email inboxes for AI agents: create an inbox, wait for the verification code, send and reply.

- Trust score: 78/100 (medium)
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-11-21

## Components

- remote · `mcp.runearc.com`: 78/100 (this document), [markdown](https://verifymcp.io/servers/runearc-labs-runearc/mcp.md), [page](https://verifymcp.io/servers/runearc-labs-runearc/mcp)

## Channel facts

- Endpoint: `https://mcp.runearc.com/mcp`
- Transports: `streamable-http`
- Auth: `required (not validated)`
- Version: `1.0.0`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically against the live server, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-11-21.

- **Endpoint Security**: 86/100
  - The endpoint's TLS certificate is valid, in date, and uses a strong key.
  - Authorisation is enforced on tool calls, but the challenge carries no valid RFC 9728 metadata, so a client cannot discover where to get a token.
  - HTTPS is enforced; there's no plaintext access path.
  - HSTS check failed: the Strict-Transport-Security header is absent.
  - DNSSEC is configured correctly; the domain's records validate against the full chain to the root.
- **Transport & Reachability**: 100/100
  - Verified streamable-http transport via a live MCP handshake.
- **Schema Quality & AI Usability**: 73/100
  - AI-judged instruction clarity (excellent).
  - Context-footprint check failed: tool/resource definitions use about 11699 tokens (~129/item across 90 items; 90 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 17/100
  - Stability observed for 5 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 97/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 91% of tool parameters carry a description.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - All 10 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation.
  - An AI judge read all 91 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.

## Install

### How do I install the Runearc MCP server?

Runearc is a hosted endpoint at https://mcp.runearc.com/mcp, so there is nothing to install locally. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add --transport http runearc-labs-runearc 'https://mcp.runearc.com/mcp'
```

### Cursor

```json
{
  "mcpServers": {
    "runearc-labs-runearc": {
      "url": "https://mcp.runearc.com/mcp"
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "runearc-labs-runearc": {
      "type": "http",
      "url": "https://mcp.runearc.com/mcp"
    }
  }
}
```

### Codex

```toml
[mcp_servers.runearc-labs-runearc]
url = "https://mcp.runearc.com/mcp"
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "runearc-labs-runearc": {
      "type": "remote",
      "url": "https://mcp.runearc.com/mcp",
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add runearc-labs-runearc --url 'https://mcp.runearc.com/mcp' --transport streamable-http
```

### Hermes

```yaml
mcp_servers:
  runearc-labs-runearc:
    url: "https://mcp.runearc.com/mcp"
```

### Netclaw

```json
{
  "McpServers": {
    "runearc-labs-runearc": {
      "Transport": "http",
      "Url": "https://mcp.runearc.com/mcp"
    }
  }
}
```

### Vellum

```bash
assistant mcp add runearc-labs-runearc -t streamable-http -u 'https://mcp.runearc.com/mcp'
```

### Other

```json
{
  "mcpServers": {
    "runearc-labs-runearc": {
      "type": "http",
      "url": "https://mcp.runearc.com/mcp"
    }
  }
}
```

The mcpServers block is a cross-client convention. Remote transports vary, so check your client's docs.

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-11-21 (score 78, +1)

No change was recorded against any check on this day. Stability & Change Management went from 13 to 17. That category is still filling its 30-day observation window: 4 days of observed history at the previous scan, 5 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-19 (score 77, +1)

No change was recorded against any check on this day. Stability & Change Management went from 7 to 10. That category is still filling its 30-day observation window: 2 days of observed history at the previous scan, 3 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-17 (score 76, +2)

- [security] Tool “add_contact” rewrote its description, which is the text the model reads
- [security] Tool “create_list” rewrote its description, which is the text the model reads
- [security] Tool “delete_segment” rewrote its description, which is the text the model reads
- [security] Tool “delete_sync_rule” rewrote its description, which is the text the model reads
- [security] Tool “goal_stats” rewrote its description, which is the text the model reads
- [security] Tool “list_event_definitions” rewrote its description, which is the text the model reads
- [security] Tool “list_lists” rewrote its description, which is the text the model reads
- [security] Tool “popup_stats” rewrote its description, which is the text the model reads
- [security] Tool “bulk_contacts” is now declared destructive
- [security] Tool “cancel_campaign” is now declared destructive
- [security] Tool “delete_inbox” is now declared destructive
- [security] Tool “delete_page” is now declared destructive
- [security] Tool “delete_segment” is now declared destructive
- [security] Tool “delete_sync_rule” is now declared destructive
- [security] Tool “remove_billing_integration” is now declared destructive
- [security] Tool “unpublish_page” is now declared destructive
- [security] Tool “unsubscribe_contact” is now declared destructive
- [security] Tool “unsubscribe_contacts” is now declared destructive
- [functional improvement] Stability: unverified → 0.03
- [functional] Destructive annotations: All 10 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation.
- [cosmetic] Tool “activate_sequence” changed its title: Activate sequence
- [cosmetic] Tool “add_contact” changed its title: Add contact
- [cosmetic] Tool “add_domain” changed its title: Add sending domain
- [cosmetic] Tool “add_to_list” changed its title: Add contacts to list
- [cosmetic] Tool “bulk_contacts” changed its title: Bulk update contacts
- [cosmetic] Tool “cancel_campaign” changed its title: Cancel campaign
- [cosmetic] Tool “connect_billing_integration” changed its title: Connect billing integration
- [cosmetic] Tool “create_campaign” changed its title: Create campaign
- [cosmetic] Tool “create_event_definition” changed its title: Create event definition
- [cosmetic] Tool “create_form” changed its title: Create signup form
- [cosmetic] Tool “create_goal” changed its title: Create goal
- [cosmetic] Tool “create_inbox” changed its title: Create inbox
- [cosmetic] Tool “create_list” changed its title: Create contact list
- [cosmetic] Tool “create_page” changed its title: Create page
- [cosmetic] Tool “create_popup” changed its title: Create popup
- [cosmetic] Tool “create_segment” changed its title: Create segment
- [cosmetic] Tool “create_sequence” changed its title: Create sequence
- [cosmetic] Tool “create_sync_rule” changed its title: Create sync rule
- [cosmetic] Tool “create_template” changed its title: Create template
- [cosmetic] Tool “create_webhook” changed its title: Create webhook
- [cosmetic] Tool “delete_inbox” changed its title: Delete inbox
- [cosmetic] Tool “delete_page” changed its title: Delete page
- [cosmetic] Tool “delete_segment” changed its title: Delete segment
- [cosmetic] Tool “delete_sync_rule” changed its title: Delete sync rule
- [cosmetic] Tool “draft_welcome_sequence” changed its title: Draft welcome sequence
- [cosmetic] Tool “enroll_contact_in_sequence” changed its title: Enroll contact in sequence
- [cosmetic] Tool “generate_email” changed its title: Generate email with AI
- [cosmetic] Tool “generate_sequence” changed its title: Generate sequence with AI
- [cosmetic] Tool “get_ai_context” changed its title: Get AI context
- [cosmetic] Tool “get_analytics” changed its title: Get email analytics
- [cosmetic] Tool “get_email_design” changed its title: Get email design
- [cosmetic] Tool “get_email_tracking” changed its title: Get email tracking
- [cosmetic] Tool “get_import” changed its title: Get import status
- [cosmetic] Tool “get_inbox” changed its title: Get inbox
- [cosmetic] Tool “get_message” changed its title: Get message
- [cosmetic] Tool “get_page” changed its title: Get page
- [cosmetic] Tool “get_sequence” changed its title: Get sequence
- [cosmetic] Tool “get_thread” changed its title: Get thread
- [cosmetic] Tool “get_tracking_settings” changed its title: Get tracking settings
- [cosmetic] Tool “goal_stats” changed its title: Get goal stats
- [cosmetic] Tool “import_contacts” changed its title: Import contacts from CSV
- [cosmetic] Tool “import_site” changed its title: Import brand from website
- [cosmetic] Tool “install_starter_sequence” changed its title: Install starter sequence
- [cosmetic] Tool “list_billing_integrations” changed its title: List billing integrations
- [cosmetic] Tool “list_campaign_designs” changed its title: List campaign designs
- [cosmetic] Tool “list_campaigns” changed its title: List campaigns
- [cosmetic] Tool “list_contact_events” changed its title: List contact events
- [cosmetic] Tool “list_contacts” changed its title: List contacts
- [cosmetic] Tool “list_domains” changed its title: List domains
- [cosmetic] Tool “list_event_definitions” changed its title: List event definitions
- [cosmetic] Tool “list_form_submissions” changed its title: List form submissions
- [cosmetic] Tool “list_forms” changed its title: List signup forms
- [cosmetic] Tool “list_goals” changed its title: List goals
- [cosmetic] Tool “list_inboxes” changed its title: List inboxes
- [cosmetic] Tool “list_lists” changed its title: List contact lists
- [cosmetic] Tool “list_messages” changed its title: List messages
- [cosmetic] Tool “list_pages” changed its title: List pages
- [cosmetic] Tool “list_segments” changed its title: List segments
- [cosmetic] Tool “list_sequences” changed its title: List sequences
- [cosmetic] Tool “list_suppressions” changed its title: List suppressions
- [cosmetic] Tool “list_sync_rules” changed its title: List sync rules
- [cosmetic] Tool “list_templates” changed its title: List templates
- [cosmetic] Tool “list_threads” changed its title: List threads
- [cosmetic] Tool “pause_campaign” changed its title: Pause campaign
- [cosmetic] Tool “pause_sequence” changed its title: Pause sequence
- [cosmetic] Tool “popup_stats” changed its title: Get popup stats
- [cosmetic] Tool “preview_segment” changed its title: Preview segment
- [cosmetic] Tool “publish_page” changed its title: Publish page
- [cosmetic] Tool “remove_billing_integration” changed its title: Remove billing integration
- [cosmetic] Tool “render_email_preview” changed its title: Render email preview
- [cosmetic] Tool “reply” changed its title: Reply to message
- [cosmetic] Tool “resume_campaign” changed its title: Resume campaign
- [cosmetic] Tool “send_campaign” changed its title: Send campaign
- [cosmetic] Tool “send_campaign_test” changed its title: Send campaign test
- [cosmetic] Tool “send_email” changed its title: Send email
- [cosmetic] Tool “send_template” changed its title: Send template email
- [cosmetic] Tool “set_email_design” changed its title: Set email design
- [cosmetic] Tool “set_sequence_steps” changed its title: Set sequence steps
- [cosmetic] Tool “tag_contacts” changed its title: Tag contacts
- [cosmetic] Tool “track_event” changed its title: Track contact event
- [cosmetic] Tool “turn_on_welcome_sequence” changed its title: Turn on welcome sequence
- [cosmetic] Tool “unpublish_page” changed its title: Unpublish page
- [cosmetic] Tool “unsubscribe_contact” changed its title: Unsubscribe contact
- [cosmetic] Tool “unsubscribe_contacts” changed its title: Unsubscribe contacts
- [cosmetic] Tool “update_campaign” changed its title: Update campaign
- [cosmetic] Tool “update_contact” changed its title: Update contact
- [cosmetic] Tool “update_page” changed its title: Update page
- [cosmetic] Tool “update_tracking_settings” changed its title: Update tracking settings
- [cosmetic] Tool “verify_domain” changed its title: Verify domain DNS
- [cosmetic] Tool “wait_for_email” changed its title: Wait for email

### 2026-11-16 (score 74, +40)

- [security improvement] Transport: unverified → pass
- [security improvement] Injection markers: unverified → pass
- [security] First check of Judged manipulation: pass
- [security] Authorization: Authorisation is enforced on tool calls, but the challenge carries no valid RFC 9728 metadata, so a client cannot discover where to get a token.
- [functional improvement] MCP protocol: unverified → pass
- [functional improvement] Tool coverage: unverified → 100
- [functional] First check of Schema quality: fail
- [functional] First check of Schema quality: excellent
- [functional] First check of Schema quality: fail
- [functional] First check of Tool coverage: 91
- [functional] First check of Destructive annotations: 0

### 2026-11-15 (score 34)

First indexed and scored.

## MCP tools (90)

### `activate_sequence` (~102 tokens)

Activate sequence

Activate a draft or paused sequence so triggers enrol contacts and due steps run. Returns the sequence with status "active". Errors: sequence_empty, from_required (set a From address with a sender on a verified domain), invalid_branch, sequence_not_found. Next: enroll_contact_in_sequence to start one contact now, or wait for the trigger.

Input parameters:

- `sequence_id` (string, required): Sequence id, such as "seq_01j9...", from list_sequences or create_sequence.

### `add_contact` (~206 tokens)

Add contact

Add ONE new contact (a person to email) to the workspace's audience; tags and list_ids given here are applied at once (missing tags are created). Use update_contact to change an existing contact, import_contacts or bulk_contacts for many. Returns the contact {id, email, status, tags, list_ids, ...}. Errors: invalid_email, contact_exists (use update_contact instead), unknown_field, not_found (list). Next: track_event for this contact, or enroll_contact_in_sequence.

Input parameters:

- `email` (string, required): The contact's email address.
- `first_name` (string|null)
- `last_name` (string|null)
- `list_ids` (array): Ids (lst_...) of lists to join.
- `revenue_cents` (integer|null): Lifetime revenue in cents.
- `source` (string|null): Where they came from, such as "signup" or "import".
- `tags` (array): Tags to attach; created when missing.

### `add_domain` (~169 tokens)

Add sending domain

Add a custom domain to send and receive mail on, such as acme.com. Returns {id, name, status, provider, can_auto_configure, records: [{type, name, value, priority, required, status, fix}], next_step, existing_mx?}. Publish every required record at the DNS host. existing_mx means the domain already gets mail elsewhere: adding our MX moves it, so prefer a subdomain such as mail.acme.com. Errors: invalid_domain, reserved_domain, domain_exists, domain_taken, insufficient_scope. Next: publish the records, then call verify_domain; then create_inbox with domain set to it.

Input parameters:

- `name` (string, required): Domain name to send and receive on, such as "acme.com" or "mail.acme.com".

### `add_to_list` (~78 tokens)

Add contacts to list

Add contacts to a list. Returns {added: n} for newly added contacts. Errors: not_found (list or contact). Next: list_contacts with list_id to see members.

Input parameters:

- `emails` (array, required): Emails or ids (ctc_...) of existing contacts.
- `list_id` (string, required): List id (lst_...).

### `bulk_contacts` (~217 tokens)

Bulk update contacts

Tag, add to a list, unsubscribe or delete many contacts at once. Give exactly one of contact_ids, emails or all_matching. Returns the operation {id, status, total, ...}; it runs in the background. Errors: invalid_body, too_many_targets, invalid_tags, empty_selection, not_found (list). Next: unsubscribe_contacts for plain unsubscribes; list_suppressions to review who is blocked.

Input parameters:

- `action` (string, required): "tag", "add_to_list", "unsubscribe" or "delete".
- `all_matching`: Act on every contact matching this filter {q, status, tag, list_id}; {} means all contacts.
- `contact_ids` (array|null): Contact ids (ctc_...). Give exactly one of contact_ids, emails or all_matching.
- `emails` (array|null): Email addresses of contacts.
- `list_id` (string|null): For "add_to_list": the list id (lst_...).
- `tags` (array): For "tag": the tags to add.

### `cancel_campaign` (~55 tokens)

Cancel campaign

Cancel a campaign that has not finished. Pending recipients are dropped; mail already sent is not recalled. Errors: campaign_not_cancelable. Next: list_campaigns.

Input parameters:

- `id` (string, required): The campaign id (cmp_...).

### `connect_billing_integration` (~145 tokens)

Connect billing integration

Connect a billing provider so purchases, renewals, cancellations and refunds become events and contact revenue. Returns webhook_path (prefix with the API origin) to paste into the provider; `secret` is returned only if Runearc generated it (shown once). Errors: unknown_billing_provider, billing_already_connected. Next: send a test event from the provider, then list_billing_integrations to see last_event_at.

Input parameters:

- `provider` (string, required): stripe, paddle, lemonsqueezy, polar, revenuecat, adapty or shopify.
- `secret` (string|null): The provider's webhook signing secret. Omit to have Runearc generate one.

### `create_campaign` (~120 tokens)

Create campaign

Create a draft campaign, blank, from a design, or as a copy of another campaign. Returns the campaign {id, status:"draft", ...}. Errors: invalid_name, unknown_design, not_found. Next: update_campaign to set subject, from_address, document and audience, then send_campaign_test.

Input parameters:

- `copy_of` (string|null): Optional campaign id (cmp_...) to copy.
- `design` (string|null): Optional design slug from list_campaign_designs.
- `name` (string, required): Campaign name, for example "November product update".

### `create_event_definition` (~105 tokens)

Create event definition

Document a custom event so teammates know what it means. Returns the definition. Errors: event_definition_exists, invalid_event_name, reserved_event_name. Next: track_event to send it, or create_sync_rule.

Input parameters:

- `description` (string|null): What the event means.
- `name` (string, required): The event name, such as "plan_upgraded".
- `properties`: Documentation of the properties: [{"name": "plan", "type": "string"}]. Not enforced.

### `create_form` (~207 tokens)

Create signup form

Create an embeddable signup form from a template (newsletter, waitlist or product-updates). Returns the form with its id. Errors: unknown_template, invalid_form_config, unknown_form_field, invalid_origin, list_not_found. Next: put <div data-runearc-form="ID"></div> and the runearc-embed.js script on your site, or create_popup to show the form on a trigger.

Input parameters:

- `allowed_origins` (array): Websites allowed to submit, such as ["https://example.com"]. Empty means any.
- `double_opt_in` (boolean): Send a confirmation email first (double opt-in).
- `list_ids` (array): Add new contacts to these lists (list ids).
- `name` (string, required): A name for the form, such as "Footer signup".
- `tags` (array): Tag new contacts with these tags.
- `template` (string|null): Starting template: newsletter, waitlist or product-updates. Default newsletter.

### `create_goal` (~167 tokens)

Create goal

Create a conversion goal. Each matching event records a conversion credited to the last email sent or clicked within window_days. Returns the goal. Errors: goal_exists, invalid_window (1 to 90), invalid_event_name. Next: track_event to convert, then goal_stats.

Input parameters:

- `conditions`: Optional property conditions, same shape as for sync rules.
- `event_name` (string, required): The event that counts as a conversion.
- `name` (string, required): A unique name, such as "Upgraded to pro".
- `value_property` (string|null): Property holding the revenue of the event, such as "amount".
- `window_days` (integer|null): Attribution window in days, 1 to 90 (default 7): an email sent or clicked this long before gets the credit.

### `create_inbox` (~246 tokens)

Create inbox

Create a new email inbox that can receive (and send) mail right away. Returns {id, address, display_name, created_at}. Errors: inbox_limit_reached (plan limit; delete an inbox or upgrade), inbox_exists (pick another name), invalid_local_part, domain_not_found (custom domain not added), insufficient_scope (key limited to one inbox). Next: give `address` to the service you are signing up for, then call wait_for_email with `id` (pass after = created_at so an email that arrives early is not missed).

Input parameters:

- `display_name` (string|null): Name shown in the From header of mail sent from this inbox, such as "Acme Bot".
- `domain` (string|null): A custom domain this org added with add_domain, such as "acme.com". Leave out to use the free shared domain (<code>.runearc.email).
- `name` (string|null): The part before the @, such as "signup-bot". 1 to 64 letters, digits, dots, hyphens or underscores. Leave out for a random name like "inbox-k3x9q2".

### `create_list` (~94 tokens)

Create contact list

Create an empty, named contact list (a static group you add people to by hand; for rule-based groups use create_segment). Names are unique per workspace. Returns {id, name, description, member_count: 0}. Errors: list_exists (call list_lists and reuse it), invalid_name. Next: add_to_list with the returned id.

Input parameters:

- `description` (string|null)
- `name` (string, required)

### `create_page` (~154 tokens)

Create page

Create a draft landing page from a template (blank, waitlist, product-launch, newsletter). Returns the page with its blocks. Errors: page_slug_taken, unknown_page_template, invalid_page_form. Next: update_page to edit blocks and pick a form, then publish_page.

Input parameters:

- `form_id` (string|null): The signup form (frm_...) the signup block uses.
- `name` (string, required): A name for the page, such as "Spring launch".
- `slug` (string|null): URL slug (3 to 40 lowercase letters, digits, hyphens). Default: from the name.
- `template` (string|null): Starting template: blank, waitlist, product-launch or newsletter. Default blank.

### `create_popup` (~225 tokens)

Create popup

Create a popup that shows a form on a trigger (time, scroll or exit_intent), optionally only on some paths. Errors: invalid_popup_config, not_found (form), invalid_status. Next: set status live, then add the runearc-embed.js script with data-popup="ID" to your site; popup_stats shows conversions.

Input parameters:

- `days_between` (integer|null): Days before showing again to the same visitor. Default 7.
- `form_id` (string, required): The form to show (frm_...).
- `heading` (string, required): The popup headline.
- `name` (string, required): A name for the popup, such as "Blog exit offer".
- `only_paths` (array): Only on URLs starting with these paths, such as ["/blog"]. Empty means every page.
- `status` (string|null): draft or live. Default draft.
- `trigger` (string|null): time, scroll or exit_intent. Default time.
- `trigger_value` (integer|null): Seconds for time, percent for scroll. Default 5 and 50.

### `create_segment` (~268 tokens)

Create segment

Save a segment (a named audience filter over contacts). Returns the segment. Errors: invalid_filter (read the fix), segment_name_taken, segment_limit. Next: preview_segment, then use the id as the audience of a campaign.

Input parameters:

- `description` (string): What it is for.
- `filter` (required): The filter: {"type":"group","op":"and"|"or","children":[...]}. Conditions: {"type":"tag","tag":"vip","has":true}, {"type":"status","is":["active"]}, {"type":"list","list_id":"lst_..","member":true},…
- `name` (string, required): Segment name, unique in the workspace.

### `create_sequence` (~191 tokens)

Create sequence

Create an empty draft sequence. Nothing sends until it has steps, a From address and is activated. Returns the sequence. Errors: invalid_name, invalid_trigger (use manual, tag:<tag>, list:<id> or event:<name>). Next: set_sequence_steps, then activate_sequence. install_starter_sequence is faster for common flows.

Input parameters:

- `from` (string|null): The From address, such as "Acme <hello@acme.com>", on a verified domain. Needed before activating a sequence that sends email.
- `goal_event` (string|null): Contacts who track this event leave the sequence, such as "upgraded".
- `name` (string, required): Sequence name, 1 to 120 characters, such as "Welcome".
- `trigger` (string, required): What enrols a contact: "manual", "tag:<tag>", "list:<list id>" or "event:<event name>".

### `create_sync_rule` (~249 tokens)

Create sync rule

Create a sync rule: when event_name happens and all conditions match, apply the action to the contact. Returns the rule. Errors: invalid_argument, invalid_event_name, list_not_found, field_not_found (create the contact field first), invalid_tag. Next: track_event to try it, then get_contact or list_contact_events to see the result.

Input parameters:

- `action` (required): What to do: {"type": "add_tag", "tag": "customer"}, {"type": "remove_tag", "tag": "trial"}, {"type": "set_field", "key": "plan", "value": "pro"} or {"type": "add_to_list", "list_id": "lst_..."}.
- `conditions`: All must match: [{"path": "plan", "op": "eq", "value": "pro"}]. Ops: eq, neq, gt, gte, lt, lte, contains, exists.
- `event_name` (string, required): The event that triggers it, custom or built-in (for example "plan_upgraded" or "tag.added").
- `name` (string, required): A name for the rule, such as "Tag pro customers".

### `create_template` (~178 tokens)

Create template

Create a transactional template from one of 18 starters (password-reset, magic-link, email-verification, verification-code, welcome, receipt, invoice, order-confirmation, shipping-notification, subscription-renewal, payment-failed, trial-ending, team-invite, login-alert, account-deleted, refund-issued, feedback-request, waitlist-confirmation) or a blank one, and publish it so it can be sent. Errors: template_slug_taken, unknown_starter, invalid_template_slug. Next: send_template.

Input parameters:

- `from_starter` (string|null): A built-in starter slug such as "password-reset" or "welcome".
- `name` (string|null): Template name. Defaults to the starter name.
- `slug` (string|null): URL-safe slug used in code. Defaults to the starter slug or the slugified name.

### `create_webhook` (~217 tokens)

Create webhook

Register a webhook: Runearc POSTs signed JSON events (Standard Webhooks) to the URL, retrying on failure. Returns {id, url, event_types, inbox_id, secret, created_at}; the secret is shown only once. Errors: invalid_webhook_url (https only, http allowed for localhost), invalid_event_type, not_found (inbox). Next: store the secret and verify each delivery's webhook-signature header with it. For a single expected email, wait_for_email is simpler than a webhook.

Input parameters:

- `event_types` (array): Event types to deliver, such as ["message.received"]. Leave out for every event. Types: message.received, message.queued, message.delivered, message.deferred, message.bounced, message.complained, mes…
- `inbox_id` (string|null): Only deliver events for this inbox. Leave out for every inbox in the org.
- `url` (string, required): Where to POST events: an https:// URL (http:// only for localhost).

### `delete_inbox` (~88 tokens)

Delete inbox

Delete an inbox and all of its messages. Mail to its address is rejected afterwards. This cannot be undone. Returns {deleted: true, id}. Errors: not_found, insufficient_scope. Next: create_inbox if you need a fresh address.

Input parameters:

- `inbox_id` (string, required): Inbox id, such as "ibx_01j9...", from create_inbox or list_inboxes.

### `delete_page` (~48 tokens)

Delete page

Delete a landing page with its view counts and lead links (the contacts stay). Errors: not_found. Next: list_pages.

Input parameters:

- `id` (string, required): The page id (pg_...).

### `delete_segment` (~65 tokens)

Delete segment

Permanently delete a saved segment (the saved filter only; the contacts in it are not touched). Built-in quick filters cannot be deleted. Returns {deleted: true}. Errors: not_found, builtin_segment_readonly. Next: list_segments.

Input parameters:

- `segment_id` (string, required)

### `delete_sync_rule` (~79 tokens)

Delete sync rule

Delete a sync rule so future events no longer trigger it. Changes it already made to contacts (tags, fields, lists) stay. Returns {deleted: id}. Errors: not_found (already deleted or wrong id; list_sync_rules shows the current ones). Next: list_sync_rules.

Input parameters:

- `id` (string, required): The id of the rule.

### `draft_welcome_sequence` (~177 tokens)

Draft welcome sequence

Draft a 4 email welcome sequence (immediately, day 1, day 3, day 5) from an imported site. It is saved as a DRAFT: nothing is sent and nobody is enrolled. Returns {sequence_id, steps:[{position, delay_seconds, subject}], address_required}. Errors: import_not_found, welcome_draft_shape. Next: if address_required is true, set the business mailing address (PUT /v1/workspace/postal-address), then call turn_on_welcome_sequence with `sequence_id` once the user agrees.

Input parameters:

- `import_id` (string, required): The import id from import_site, such as "sim_01j9...".
- `style` (string|null): "designed" (branded layout, default) or "plain" (looks like a personal email).

### `enroll_contact_in_sequence` (~176 tokens)

Enroll contact in sequence

Enrol one contact in an active sequence now. Returns the enrolment {id, status, current_position, next_run_at}. Errors: sequence_not_active, contact_not_found, contact_not_mailable (unsubscribed or suppressed), already_enrolled, reenroll_not_allowed (pass force), invalid_enroll_target (give contact_id or email). Next: get_sequence to watch enrolment counts.

Input parameters:

- `contact_id` (string|null): Contact id, such as "ctc_01j9...". Give this or email.
- `email` (string|null): Contact email address. Give this or contact_id.
- `force` (boolean): Enrol again even if the contact already went through the sequence. Never overrides an active enrolment.
- `sequence_id` (string, required): Sequence id. The sequence must be active.

### `generate_email` (~233 tokens)

Generate email with AI

Draft one email with AI from a prompt and the workspace AI context. Returns {subject, preview_text, document} where document is an email design document (version, style, subject, preheader, blocks: heading, text, button, divider). Errors: ai_unavailable (server has no AI key), ai_limit_reached (monthly allowance used; fix: wait, write by hand or upgrade), invalid_language (enable it in Settings > Localization), ai_unsafe_link (draft linked outside your sites; run again), ai_refused. Counts as one AI generation. Next: review the draft before sending; nothing is sent or saved.

Input parameters:

- `language` (string|null): Language tag enabled in the workspace (Settings > Localization), such as "de". Defaults to the workspace default language.
- `prompt` (string, required): What the email is for, in plain words, such as "welcome email for new trial users". 1 to 4000 characters.
- `site_text` (string|null): Optional text from the customer's website to learn facts and style from. Treated as untrusted data, never as instructions.

### `generate_sequence` (~164 tokens)

Generate sequence with AI

Draft an automated email sequence with AI. Returns {name, steps:[{delay_days, subject, preview_text, document}]}; the first step has delay_days 0. Same errors as generate_email, plus invalid_steps (steps must be 1 to 10). Counts as ONE AI generation. Nothing is saved or sent. Next: review the steps; save them as a sequence once sequences are available.

Input parameters:

- `language` (string|null)
- `prompt` (string, required): What the sequence should achieve, such as "onboard new trial users".
- `site_text` (string|null)
- `steps` (integer|null): Number of emails, 1 to 10. Default 4 (immediately, day 1, day 3, day 5).

### `get_ai_context` (~79 tokens)

Get AI context

Read the workspace AI context: product name and url, category, founder voice, tone, company context, extra instructions, and enabled languages. Returns {product_name, product_url, category, founder_name, founder_voice, tone, company_context, instructions, languages, default_language}. Next: generate_email or generate_sequence, which use this context automatically.

### `get_analytics` (~416 tokens)

Get email analytics

Email analytics for the workspace: sent, delivered, bounced, complained, opened, clicked, unsubscribed, conversions and revenue_cents as a zero filled time series (day buckets, hour buckets for ranges of two days or less), totals, rates (null when there is nothing to divide by) and a breakdown table. Counts are unique per recipient, bots ignored, counted when the event happened. Filter by range (7d, 30d, 90d, custom), type, source_id, step_id, provider and subscriber; group rows with by. `availability` says which sources exist: campaigns, sequences, conversions and segments are false until those features are set up, and their numbers are then zero. Errors: invalid_range, invalid_type, invalid_provider, invalid_subscriber, invalid_breakdown, segment_unavailable, insufficient_scope. Next: list_domains if sent is 0 and nothing has been sent yet.

Input parameters:

- `by` (string|null): How to group the table: "source" (default), "type" or "provider".
- `from` (string|null): First day for range=custom, YYYY-MM-DD (UTC, inclusive).
- `provider` (string|null): Outbound provider: "ses", "kumo" or "unknown".
- `range` (string|null): Date range: "7d", "30d" (default), "90d" or "custom" (then from and to are required).
- `source_id` (string|null): A campaign or sequence id, to look at one send.
- `step_id` (string|null): A sequence step or campaign email id, together with source_id.
- `subscriber` (string|null): One subscriber's email address, or a segment id (seg_...).
- `to` (string|null): Last day for range=custom, YYYY-MM-DD (UTC, inclusive). At most 366 days in total.
- `type` (string|null): Send type: "transactional", "campaign" or "sequence".

### `get_email_design` (~72 tokens)

Get email design

Get the workspace email design: brand color, font, alignment, logo, header, footer and postal address. Returns {brand, design_md}; design_md is a markdown file you can edit and send back with set_email_design. Next: set_email_design, or render_email_preview to see how a document looks.

### `get_email_tracking` (~120 tokens)

Get email tracking

Counted opens, unique opens, clicks, unique clickers and unsubscribes of one email, plus how many opens and clicks the bot filter flagged. Returns {opens, unique_opens, clicks, unique_clickers, unsubscribes, flagged_opens, flagged_clicks}. Errors: forbidden. Next: list the email's events with GET /v1/tracking/events?message_id=..., or change what is tracked with update_tracking_settings.

Input parameters:

- `email_id` (string, required): Email id such as "eml_01j9...".

### `get_import` (~109 tokens)

Get import status

Progress of a contact import. Returns {import: {status, total_rows, processed_rows, created_count, updated_count, skipped_count, error_count}, errors: [{row_number, email, code, message}]} (first 50 errors). Errors: not_found. Next: call again while status is queued or running; fix the rows in `errors` and import them again.

Input parameters:

- `import_id` (string, required): Import id, such as "imp_01j9...", from import_contacts.

### `get_inbox` (~70 tokens)

Get inbox

Get one inbox by id. Returns {id, address, display_name, created_at}. Errors: not_found. Next: list_messages or wait_for_email.

Input parameters:

- `inbox_id` (string, required): Inbox id, such as "ibx_01j9...", from create_inbox or list_inboxes.

### `get_message` (~150 tokens)

Get message

Get one message with its text body and the one time codes and links found in it. Returns {id, inbox_id, thread_id, direction, status, from, to, cc, subject, text, codes, links, attachments, spam, created_at}; html only when include_html is true. Errors: not_found. Next: use codes[0] or a link from links to finish a sign up, or call reply to answer.

Input parameters:

- `include_html` (boolean): Also return the HTML body. Off by default because the text body, codes and links are usually all an agent needs.
- `message_id` (string, required): Message id, such as "msg_01j9...".

### `get_page` (~54 tokens)

Get page

Read one landing page: draft blocks, status, form and public url. Errors: not_found. Next: update_page to edit it, or publish_page.

Input parameters:

- `id` (string, required): The page id (pg_...).

### `get_sequence` (~77 tokens)

Get sequence

Get one sequence with its trigger, goal, From address, ordered steps and enrolment counts. Returns the sequence. Errors: sequence_not_found. Next: set_sequence_steps to edit, activate_sequence to start it.

Input parameters:

- `sequence_id` (string, required): Sequence id, such as "seq_01j9...", from list_sequences or create_sequence.

### `get_thread` (~97 tokens)

Get thread

Get a conversation with all of its messages, oldest first, each with text, codes and links. Returns {id, inbox_id, subject, message_count, last_message_at, messages: [...]}. Errors: not_found. Next: reply with the id of the last message to continue the conversation.

Input parameters:

- `thread_id` (string, required): Thread id, such as "thr_01j9...", from list_threads or a message's thread_id.

### `get_tracking_settings` (~66 tokens)

Get tracking settings

Show the workspace's tracking settings: opens, clicks, unsubscribe, strict bot filter and the per-address frequency cap. Returns {track_opens, track_clicks, unsubscribe_enabled, strict_bot_filter, frequency_cap_count, frequency_cap_hours}. Next: update_tracking_settings to change them.

### `goal_stats` (~106 tokens)

Get goal stats

Read-only results of one goal over the last `days` (default 30, 1 to 365): conversions, unique contacts, how many were credited to an email, revenue_cents and the top emails. Errors: not_found. Next: list_goals, or create_goal for another outcome.

Input parameters:

- `days` (integer|null): Look back this many days (1 to 365, default 30).
- `id` (string, required): The goal id (gol_...).

### `import_contacts` (~223 tokens)

Import contacts from CSV

Import contacts from CSV text and start the import. Detects the email, first name and last name columns from the header. Returns {import, mapping_used}; the import runs in the background. Errors: empty_csv, invalid_csv, csv_too_large, too_many_rows, unknown_column, not_found (list). Next: get_import with import.id until import.status is done; rows that failed or were skipped are in `errors`.

Input parameters:

- `csv` (string, required): The whole CSV file as text. First row is the header. Up to 25 MB.
- `email_column` (string|null): Header of the email column. Leave out to detect it ("email", "e-mail", ...).
- `filename` (string|null): File name for the record, such as "customers.csv".
- `list_ids` (array): List ids (lst_...) every imported contact is added to.
- `tags` (array): Tags to add to every imported contact.
- `update_existing` (boolean): Merge into contacts that already exist instead of skipping them. Default false.

### `import_site` (~125 tokens)

Import brand from website

Read a website and save its name, logo, brand colour and description as the workspace's brand and company context. Fetches only public web pages (private and internal addresses are refused). Returns {id, name, description, logo_url, color}. Errors: site_blocked, invalid_site_url, site_unreachable, site_not_html, too_many_redirects, site_import_rate_limited. Next: call draft_welcome_sequence with `id`.

Input parameters:

- `url` (string, required): Public address of the website, such as "acme.com" or "https://acme.com".

### `install_starter_sequence` (~97 tokens)

Install starter sequence

Create a draft sequence from a ready made one: welcome, onboarding, trial-ending, win-back, churn-save or review-request, with real email copy. Returns the sequence. Errors: starter_not_found. Next: update the From address and button links, then activate_sequence.

Input parameters:

- `key` (string, required): One of "welcome", "onboarding", "trial-ending", "win-back", "churn-save", "review-request".

### `list_billing_integrations` (~63 tokens)

List billing integrations

List connected billing providers (Stripe, Paddle, Lemon Squeezy, Polar, ...). Returns {integrations: [...]} with each webhook_path (prefix with the API origin) and last_event_at. Next: connect_billing_integration to add one.

### `list_campaign_designs` (~39 tokens)

List campaign designs

List the 8 starting designs for a campaign (slug, name, subject, document). Next: create_campaign with design set to a slug.

### `list_campaigns` (~91 tokens)

List campaigns

List broadcast campaigns, newest first, optionally only one status (draft, scheduled, sending, paused, sent, canceled). Each has counters: recipients_total, sent_count, suppressed_count, deferred_count, unsubscribed_count, failed_count. Next: create_campaign, or send_campaign for a draft.

Input parameters:

- `status` (string|null): Only this status: draft, scheduled, sending, paused, sent or canceled.

### `list_contact_events` (~118 tokens)

List contact events

List a contact's events, newest first. Returns {data: [...], next_cursor}. Errors: not_found (contact). Next: track_event to add one, or pass next_cursor back as cursor for the next page.

Input parameters:

- `contact_id` (string, required): The contact id (ctc_...).
- `cursor` (string|null): The next_cursor of the previous page.
- `limit` (integer|null): How many events (1 to 100, default 20).
- `name` (string|null): Only events with this name.

### `list_contacts` (~140 tokens)

List contacts

Find contacts, newest first. Filter by q, status, tag or list_id. Returns {contacts: [...], next_cursor}. Pass next_cursor back as cursor for the next page. Next: update_contact, tag_contacts or unsubscribe_contact on a result.

Input parameters:

- `cursor` (string|null): next_cursor from the previous page.
- `limit` (integer|null): 1 to 200, default 50.
- `list_id` (string|null)
- `q` (string|null): Substring of email, first or last name.
- `status` (string|null): active, unsubscribed, bounced or complained.
- `tag` (string|null)

### `list_domains` (~42 tokens)

List domains

List the org's custom domains with their status and DNS records. Returns {domains: [...]}. Next: verify_domain for any that are not verified, or add_domain.

### `list_event_definitions` (~76 tokens)

List event definitions

List the custom event names this workspace has documented (name, description, property hints), so you use the same names in track_event and create_sync_rule. Read-only; events can be tracked without a definition. Returns {events: [...]}. Next: create_event_definition to document another, or create_sync_rule to act on one.

### `list_form_submissions` (~104 tokens)

List form submissions

Recent submissions of a form, newest first: {submissions:[{email,status,origin,page_url,created_at}]}. Status is accepted, pending_confirmation, confirmed, suppressed or spam. Errors: not_found (form). Next: list_contacts to see the people who joined, or create_popup to collect more signups.

Input parameters:

- `form_id` (string, required): The form id (frm_...).
- `limit` (integer|null): Max rows, default 50.

### `list_forms` (~58 tokens)

List signup forms

List the workspace's signup forms. Returns {forms:[{id,name,status,config,...}]}. Errors: insufficient_scope (needs a full access key). Next: list_form_submissions for who signed up, or create_form for a new one.

### `list_goals` (~41 tokens)

List goals

List conversion goals (an event plus conditions that counts as a win). Returns {goals: [...]}. Next: goal_stats for results, or create_goal.

### `list_inboxes` (~58 tokens)

List inboxes

List the inboxes this API key can use. Returns {inboxes: [{id, address, display_name, created_at}]}. Next: list_messages or wait_for_email with an inbox id, or create_inbox if there is none.

### `list_lists` (~60 tokens)

List contact lists

List every contact list in the workspace (static lists, not segments) with member counts. Read-only. Returns {lists: [{id, name, description, member_count}]}. Next: add_to_list, or list_contacts with list_id to see members.

### `list_messages` (~172 tokens)

List messages

List messages in an inbox, newest first, received and sent. Returns {messages: [{id, thread_id, direction, from, to, subject, preview, attachments, created_at}], next_cursor}. Pass next_cursor back as cursor for older messages. Errors: not_found (inbox). Next: get_message for the full text, codes and links; reply to answer one. To wait for a message that has not arrived yet, use wait_for_email instead of polling this.

Input parameters:

- `cursor` (string|null): The next_cursor from the previous page, to get older messages.
- `inbox_id` (string, required): Inbox id, such as "ibx_01j9...".
- `limit` (integer|null): How many messages to return, newest first. 1 to 100, default 20.

### `list_pages` (~54 tokens)

List pages

List the workspace's landing pages with status, views, leads and conversion per view. Returns {pages: [...]}. Errors: insufficient_scope (needs a full access key). Next: create_page, or publish_page for a draft.

### `list_segments` (~50 tokens)

List segments

List segments: six built-in quick filters (builtin_*) then saved ones, each with a filter and a cached last_count. Next: preview_segment for a live count, create_segment to save your own.

### `list_sequences` (~93 tokens)

List sequences

List sequences (lifecycle automations), newest first. Returns {sequences: [{id, name, status, trigger, steps, enrolled, active, completed, ...}]}. Next: get_sequence for one, install_starter_sequence for a ready made one, or create_sequence.

Input parameters:

- `status` (string|null): Only sequences with this status: "draft", "active" or "paused". Leave out for all.

### `list_suppressions` (~121 tokens)

List suppressions

List addresses that will not be emailed: hard bounces, complaints, manual blocks and unsubscribed contacts. Returns {suppressions: [{address, reason, created_at}], next_cursor, counts}. Errors: invalid_reason, invalid_cursor. Next: pass next_cursor as cursor for more.

Input parameters:

- `cursor` (string|null): next_cursor of the previous page.
- `q` (string|null): Only addresses starting with this text.
- `reason` (string|null): "hard_bounce", "complaint", "manual" or "unsubscribed".

### `list_sync_rules` (~48 tokens)

List sync rules

List sync rules: when an event matches, tag the contact, set a field or add it to a list. Returns {rules: [...]}. Next: create_sync_rule or delete_sync_rule.

### `list_templates` (~43 tokens)

List templates

List this workspace's transactional templates. Returns {data:[{id, slug, name, published_version_id}]}. Next: send_template, or create_template to add one.

### `list_threads` (~93 tokens)

List threads

List conversations (threads) in an inbox, most recently active first, up to 100. Returns {threads: [{id, inbox_id, subject, message_count, last_message_at}]}. Errors: not_found (inbox). Next: get_thread to read a conversation.

Input parameters:

- `inbox_id` (string, required): Inbox id, such as "ibx_01j9...", from create_inbox or list_inboxes.

### `pause_campaign` (~48 tokens)

Pause campaign

Pause a sending campaign; pending recipients stay queued. Errors: campaign_not_pausable. Next: resume_campaign or cancel_campaign.

Input parameters:

- `id` (string, required): The campaign id (cmp_...).

### `pause_sequence` (~86 tokens)

Pause sequence

Pause an active sequence: no new contacts enter and no step runs; enrolments keep their place and catch up on resume. Returns the sequence with status "paused". Errors: sequence_not_found. Next: set_sequence_steps to edit it, then activate_sequence.

Input parameters:

- `sequence_id` (string, required): Sequence id, such as "seq_01j9...", from list_sequences or create_sequence.

### `popup_stats` (~107 tokens)

Get popup stats

Read-only performance of one popup made with create_popup: impressions, dismissals, conversions and conversion_rate (conversions / impressions, 0 when there are no impressions). Returns {impressions, dismissals, conversions, conversion_rate}. Errors: not_found (popup; use the id create_popup returned). Next: change the trigger or paths if the rate is low, or list_form_submissions to see who converted.

Input parameters:

- `popup_id` (string, required): The popup id (pop_...).

### `preview_segment` (~126 tokens)

Preview segment

Live count and a sample of contacts for a segment id or a draft filter, without saving. Returns {count, sample:[{id,email,first_name,last_name,status}]}. Engagement and event conditions match nobody until tracking data exists. Next: create_segment to save it.

Input parameters:

- `filter`: A draft filter, same shape as create_segment.
- `limit` (integer|null): Sample size, 1 to 100. Default 10.
- `segment_id` (string|null): A saved or built-in segment id (seg_..., builtin_engaged). Give this or `filter`.

### `publish_page` (~73 tokens)

Publish page

Publish a landing page: freezes the draft as the live snapshot at /p/<workspace>/<slug>. Errors: page_needs_form (a signup block needs a form_id). Next: share the returned url, or list_pages to watch views and leads.

Input parameters:

- `id` (string, required): The page id (pg_...).

### `remove_billing_integration` (~55 tokens)

Remove billing integration

Disconnect a billing provider; its webhook URL stops working. Returns {deleted}. Errors: not_found. Next: list_billing_integrations to confirm.

Input parameters:

- `id` (string, required): The connection id (bil_...).

### `render_email_preview` (~194 tokens)

Render email preview

Render an email document with the workspace brand, without sending. Returns {subject, html, text}. Errors: invalid_document, unknown_variable (add a fallback like {{FIRST_NAME|Friend}} or pass vars), missing_address, invalid_variable. Next: send it, or fix the document and render again.

Input parameters:

- `document` (required): The email document: {version:1, style:"designed"|"plain", subject, preheader?, blocks:[{type:"heading",text,level}|{type:"text",text}|{type:"button",label,url}| {type:"image",src,alt,href?}|{type:"di…
- `vars` (object|null): Values for {{NAME}} variables, such as {"FIRST_NAME":"Ana"}.

### `reply` (~189 tokens)

Reply to message

Reply to a message in the same thread (sets In-Reply-To, References and a "Re:" subject). Replying to a message you sent follows up with its recipients. Returns {id, thread_id, status, from, to, cc, subject, created_at}. Errors: not_found (message), missing_body, no_reply_address, recipient_suppressed, daily_send_limit_reached, sending_paused, transport_unavailable. Next: wait_for_email on the inbox for their answer.

Input parameters:

- `html` (string|null): HTML body. Give text, html or both.
- `message_id` (string, required): Id of the message to reply to, such as "msg_01j9...".
- `reply_all` (boolean): Also copy everyone else who was on the original message. Default false.
- `text` (string|null): Plain text body. Give text, html or both.

### `resume_campaign` (~52 tokens)

Resume campaign

Resume a paused campaign; sending continues with the pending recipients. Errors: campaign_not_resumable. Next: list_campaigns to watch the counters.

Input parameters:

- `id` (string, required): The campaign id (cmp_...).

### `send_campaign` (~126 tokens)

Send campaign

Send a draft campaign now, or schedule it with scheduled_at (RFC 3339, within 30 days). Runearc sends one personalised email per active contact, skipping unsubscribed and suppressed ones and spacing out contacts over the frequency cap. Errors: campaign_incomplete, campaign_not_sendable, invalid_schedule. Next: list_campaigns to watch the counters, pause_campaign or cancel_campaign to stop.

Input parameters:

- `id` (string, required): The campaign id (cmp_...).
- `scheduled_at` (string|null): RFC 3339 time to send at; leave out to send now.

### `send_campaign_test` (~86 tokens)

Send campaign test

Send a campaign test to 1 to 5 addresses (subject prefixed [Test], no tracking, no recipient rows). Errors: campaign_incomplete (read the fix), invalid_test_recipients, campaign_canceled. Next: send_campaign.

Input parameters:

- `id` (string, required): The campaign id (cmp_...).
- `to` (array, required): 1 to 5 addresses.

### `send_email` (~238 tokens)

Send email

Send a new email from an inbox. Starts a new thread. Returns {id, thread_id, status, from, to, cc, subject, created_at}; status is "queued" until delivery. Errors: missing_body (give text or html), missing_recipients, invalid_address, invalid_subject, too_many_recipients, recipient_suppressed (address bounced or complained before), daily_send_limit_reached, monthly_send_limit_reached, sending_paused, transport_unavailable (sending not enabled on this server), not_found (inbox). Next: wait_for_email on the same inbox to catch the answer, or create_webhook for message.delivered events.

Input parameters:

- `cc` (array): Copied recipient addresses.
- `html` (string|null): HTML body. Give text, html or both.
- `inbox_id` (string, required): Inbox id to send from, such as "ibx_01j9...". The From address is the inbox address.
- `subject` (string, required)
- `text` (string|null): Plain text body. Give text, html or both.
- `to` (array, required): Recipient addresses.

### `send_template` (~161 tokens)

Send template email

Send an email from a published transactional template. Returns {id}. Errors: template_not_found, template_not_published, missing_variable (names the variable), invalid_variable, template_conflict, recipient_suppressed, daily_send_limit_reached, sending_paused. Next: wait_for_email or create_webhook for delivery events.

Input parameters:

- `from` (string, required): Sender on a verified domain, such as "Acme <noreply@acme.com>".
- `subject` (string|null): Overrides the template subject.
- `template` (string, required): Template slug or tpl_ id. Must have a published version.
- `to` (array, required)
- `variables`: Values for the template variables, for example {"RESET_URL": "https://..."}.

### `set_email_design` (~114 tokens)

Set email design

Replace the workspace email design from a design.md file. Returns {brand}. Errors: invalid_design_md, invalid_brand (each has a fix), forbidden (needs a full scope key). Marketing email needs the ## Address section (CAN-SPAM). Next: render_email_preview.

Input parameters:

- `design_md` (string, required): The whole design.md file: front matter (color, font, align, logo) then optional ## Header, ## Footer and ## Address sections. Get the current one with get_email_design and edit it.

### `set_sequence_steps` (~207 tokens)

Set sequence steps

Replace all steps of a draft or paused sequence. Returns the sequence. Errors: sequence_active (pause it first), invalid_steps (the steps JSON does not parse), sequence_empty, too_many_steps, invalid_delay, invalid_email_step, invalid_branch (branches jump forward only). Next: activate_sequence.

Input parameters:

- `sequence_id` (string, required): Sequence id. The sequence must be a draft or paused.
- `steps` (array, required): The ordered steps. Each is {"kind":"email","subject":"..","preview":"..","doc":{"version":1,"style":"designed","subject":"..","blocks":[{"type":"text","text":"Hi {{FIRST_NAME|there}}"}]}}, {"kind":"d…

### `tag_contacts` (~84 tokens)

Tag contacts

Attach a tag to contacts (the tag is created if missing). Returns {tagged: n} for newly tagged contacts. Errors: not_found (contact), invalid_tag. Next: list_contacts with tag to see them.

Input parameters:

- `emails` (array, required): Emails or ids (ctc_...) of existing contacts.
- `tag` (string, required): Tag name; created when missing.

### `track_event` (~227 tokens)

Track contact event

Record that a contact did something, such as plan_upgraded. Runearc then applies sync rules (tags, fields, lists) and goals. Returns the event with duplicate=true when the idempotency_key was seen before. Errors: contact_not_found (create the contact first), reserved_event_name, invalid_event_name, invalid_properties, properties_too_large, idempotency_key_reused. Next: list_contact_events to confirm it, or create_sync_rule to react to it.

Input parameters:

- `email` (string, required): Email of an existing contact the event is about.
- `idempotency_key` (string|null): Repeating the same key returns the first event instead of recording a second one.
- `name` (string, required): The event name: lowercase letters, digits and _ . : - starting with a letter, such as "plan_upgraded". Built-in names (email.opened, tag.added, ...) are reserved.
- `properties`: Event properties as a JSON object, up to 16 KB, such as {"plan": "pro", "amount": 49}.

### `turn_on_welcome_sequence` (~109 tokens)

Turn on welcome sequence

Turn on the welcome sequence: from now on contacts matching its audience are enrolled and emailed. Ask the user first, this is the step that sends mail. Returns {sequence_id, active: true}. Errors: postal_address_required (CAN-SPAM, add the address first), welcome_not_found. Next: add contacts, or watch the sequence's analytics.

Input parameters:

- `sequence_id` (string, required): The welcome sequence id from draft_welcome_sequence, such as "seq_01j9...".

### `unpublish_page` (~55 tokens)

Unpublish page

Take a landing page offline; the address answers not found. Counts and leads are kept. Errors: not_found. Next: publish_page to bring it back.

Input parameters:

- `id` (string, required): The page id (pg_...).

### `unsubscribe_contact` (~59 tokens)

Unsubscribe contact

Unsubscribe a contact so broadcasts skip them. Returns the contact with status unsubscribed. Errors: not_found. Next: list_contacts with status unsubscribed to review.

Input parameters:

- `email` (string, required): The contact's email or id (ctc_...).

### `unsubscribe_contacts` (~93 tokens)

Unsubscribe contacts

Unsubscribe addresses from marketing mail. Unknown addresses become unsubscribed contacts, so a later import cannot resubscribe them. Transactional mail still sends. Returns the operation {id, status, total}. Errors: too_many_targets, empty_selection. Next: list_suppressions with reason "unsubscribed".

Input parameters:

- `emails` (array, required): Addresses to unsubscribe from marketing mail (max 10000). Unknown addresses are remembered.

### `update_campaign` (~223 tokens)

Update campaign

Edit a draft campaign: name, subject, preheader, from_address (a verified sender), reply_to, document (email design JSON) and audience ({kind:list,id} | {kind:segment,id} | {kind:tag,name}). Only drafts can be edited. Errors: campaign_not_editable, invalid_address, invalid_document, invalid_field, not_found. Next: send_campaign_test, then send_campaign.

Input parameters:

- `audience`: {kind:"list",id} | {kind:"segment",id} | {kind:"tag",name}.
- `document`: Email design document: {version:1, style, subject, blocks:[...]}.
- `from_address` (string|null): Sender address on a verified domain, such as news@example.com.
- `id` (string, required): The campaign id (cmp_...).
- `name` (string|null)
- `preheader` (string|null)
- `reply_to` (string|null)
- `subject` (string|null): Subject line; variables like {{FIRST_NAME|there}} are allowed.

### `update_contact` (~104 tokens)

Update contact

Change a contact's names, source or revenue. Pass an empty string to clear a name. Returns the contact. Errors: not_found, invalid_revenue. Next: list_contacts to confirm.

Input parameters:

- `email` (string, required): The contact's email or id (ctc_...).
- `first_name` (string|null): Empty string clears.
- `last_name` (string|null)
- `revenue_cents` (integer|null)
- `source` (string|null)

### `update_page` (~186 tokens)

Update page

Change a landing page draft: name, slug, SEO title, description, signup form_id (null clears it) or the whole blocks array. A published page keeps serving its snapshot until publish_page. Errors: invalid_page_link, too_many_blocks, too_many_signup_blocks, page_slug_taken, invalid_page_form. Next: publish_page.

Input parameters:

- `blocks` (array|null): The whole ordered blocks array. Types: hero, features, signup, cta, footer.
- `clear_form` (boolean): Set true to remove the signup form from the page.
- `description` (string|null): SEO description.
- `form_id` (string|null): The signup form (frm_...).
- `id` (string, required): The page id (pg_...).
- `name` (string|null)
- `slug` (string|null)
- `title` (string|null): SEO title.

### `update_tracking_settings` (~222 tokens)

Update tracking settings

Change tracking settings; only the fields you pass change. Marketing mail is mail tagged category=marketing: only it gets unsubscribe links and the frequency cap. Errors: invalid_frequency_cap (1 to 1000 emails per 1 to 8760 hours), forbidden. Returns the new settings. Next: get_tracking_settings to confirm, then send_email with the category=marketing tag.

Input parameters:

- `clear_frequency_cap` (boolean): Set true to remove the cap.
- `frequency_cap_count` (integer|null): Max marketing emails per address per window; use with frequency_cap_hours.
- `frequency_cap_hours` (integer|null): Window of the cap in hours.
- `strict_bot_filter` (boolean|null): Flag opens and clicks from scanners, prefetchers and instant hits instead of counting them.
- `track_clicks` (boolean|null): Rewrite links to signed redirects that count clicks.
- `track_opens` (boolean|null): Track opens with a first-party pixel.
- `unsubscribe_enabled` (boolean|null): Add List-Unsubscribe headers and a footer link to marketing mail.

### `verify_domain` (~122 tokens)

Verify domain DNS

Check a domain's DNS records now and update its status. Returns the domain with status "verified", "pending" or "failing", and per record status ok, missing or wrong with a fix. DNS changes can take a few minutes to be visible. Errors: domain_not_found. Next: if not verified, apply each record's fix and call this again; once verified, create_inbox with domain set.

Input parameters:

- `domain` (string, required): Domain name, such as "acme.com", or its id, such as "dom_01j9...".

### `wait_for_email` (~397 tokens)

Wait for email

Wait until an email arrives in an inbox, then return it with the one time codes and links extracted, so a sign up is two calls: create_inbox, then this. Blocks up to timeout_seconds (default 60, max 120). Returns {message: {id, from, subject, text, codes, links, ...}, codes, links}: codes are 4 to 8 digit OTPs, most likely first; links are https URLs such as verification links. Only messages received after `after` (default: now) match, so pass the inbox's created_at if the email may already be there. With the catch-all inbox (any address @{code}.runearc.email), pass `to` with the address you signed up with. Errors: wait_timeout (nothing matched in time: safe to call again, or loosen from/subject_contains), invalid_timeout, not_found (inbox). Next: enter codes[0] or open the verification link.

Input parameters:

- `after` (string|null): Only match messages received after this RFC 3339 time, such as the inbox's created_at. Default: when this call starts. Set it when the email may already have arrived.
- `from` (string|null): Only match senders whose address contains this, case insensitive, such as "github.com".
- `inbox_id` (string, required): Inbox id to watch, such as "ibx_01j9...".
- `subject_contains` (string|null): Only match subjects that contain this, case insensitive, such as "verify".
- `timeout_seconds` (integer|null): Seconds to wait, 1 to 120. Default 60.
- `to` (string|null): Only match mail sent to this exact address (envelope recipient, To or Cc), such as "signup-42@k7m3qz.runearc.email". Lets several waits share the catch-all inbox.

## Diagnostics

Captured diagnostic sections: TLS, DNSSEC, Authorisation, Transports. The full working is on the page: https://verifymcp.io/servers/runearc-labs-runearc/mcp#diagnostics

## Score history

- 2026-11-21: 78
- 2026-11-20: 77
- 2026-11-19: 77
- 2026-11-18: 76
- 2026-11-17: 76
- 2026-11-16: 74
- 2026-11-15: 34

## Common questions

### What is the Runearc MCP server?

Runearc is an MCP server listed in the public MCP registry as io.github.runearc-labs/runearc. Email inboxes for AI agents: create an inbox, wait for the verification code, send and reply. This page covers its hosted endpoint (https://mcp.runearc.com/mcp).

### Is the Runearc MCP server safe to use?

Runearc scores 78 out of 100 on VerifyMCP. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the Runearc MCP server expose?

Runearc exposes 90 tools: activate_sequence, add_contact, add_domain, add_to_list, bulk_contacts, and 85 more. Their descriptions and schemas cost roughly 11,591 tokens of context every time the server is loaded.

### Does the Runearc MCP server require authentication?

Yes. Runearc asked us for credentials when we connected, so you will need to authorise it in your MCP client before it can do anything.

### Is the Runearc MCP server still maintained?

Runearc is still listed as active in the MCP registry. We last reached this channel on 21 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

### What did VerifyMCP test on the Runearc MCP server?

VerifyMCP ran 13 checks on Runearc, covering TLS certificate, Authorization, HTTPS, HSTS header and 9 more. Supply Chain Security and Provenance & Transparency do not apply to a hosted endpoint, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.

## Links

- Remote endpoint: https://mcp.runearc.com/mcp
- Authorisation metadata: https://mcp.runearc.com/.well-known/oauth-protected-resource/mcp
- Repository: https://github.com/runearc-labs/runearc-mcp
- Website: https://runearc.com/docs/mcp
- Changelog RSS feed: https://verifymcp.io/servers/runearc-labs-runearc/mcp.xml
- Changelog JSON feed: https://verifymcp.io/servers/runearc-labs-runearc/mcp/changelog.json
- Trust report (JSON): https://verifymcp.io/servers/runearc-labs-runearc/mcp.json
- HTML version of this page: https://verifymcp.io/servers/runearc-labs-runearc/mcp
