> **Building with AI coding agents?** Install the authstack plugin with one command. This equips your agent with accurate Scalekit implementation patterns.
>
> **Recommended** (the `-y` flags skip prompts, so the command runs without an interactive terminal):
> ```bash
> npx -y @scalekit-inc/cli setup -y
> ```
>
> Global:
> ```bash
> npm install -g @scalekit-inc/cli
> scalekit setup -y
> ```
>
> Supports Claude Code, Cursor, GitHub Copilot, Codex + skills for other Agent Skills-compatible agents in `~/.agents/skills`.
> Skills: integrate-agentkit, implement-saaskit, add-mcp-oauth, implement-sso, implement-scim.
> [Full setup guide](https://docs.scalekit.com/dev-kit/build-with-ai/)

---

# Connect AI agents to the Hunter MCP server

The Hunter MCP connector routes your AI agent's tool calls to Hunter's own MCP server through Scalekit. Each user signs in to Hunter once, and Scalekit stores and refreshes their tokens, so your agent never handles credentials. It comes with 112 tools.

**Authentication:** OAuth 2.1/DCR
**Categories:** CRM & Sales, Marketing
**Tools:** 112: 43 read, 32 write, 37 destructive
**Users sign in with:** OAuth
**OAuth app:** Your own Hunter MCP server app
**Built by:** Vendor MCP
**Try it:** [Playground in the Scalekit dashboard](https://app.scalekit.com/ws/signup?sk_intent=playground&provider=HUNTERMCP)

## Setup

1. ### Install the SDK

   **Node.js**

   ```bash
   npm install @scalekit-sdk/node dotenv
   ```

   **Python**

   ```bash
   pip install scalekit-sdk-python python-dotenv
   ```

2. ### Set your credentials

   Add your Scalekit credentials to your `.env` file. Find values in **[app.scalekit.com](https://app.scalekit.com)** > **Developers** > **API Credentials**.

   ```sh title=".env"
   SCALEKIT_ENVIRONMENT_URL=<your-environment-url>
   SCALEKIT_CLIENT_ID=<your-client-id>
   SCALEKIT_CLIENT_SECRET=<your-client-secret>
   ```

3. ### Create the Hunter MCP connection

   In **AgentKit > Connections**, create a Hunter MCP connection and copy its redirect URI. The name you give it is the `connection_name` your code passes. See [Configure connections](/agentkit/connections/).

4. ### Register an OAuth app

   Hunter MCP server connections use your own OAuth app. Register one with Hunter MCP server and add the redirect URI you copied.

   Then enter the app's Client ID and Client Secret on the Hunter MCP connection.

5. ### Authorize a user and make your first call

   **Node.js** (`quickstart.mts`)

   ```typescript
   import { ScalekitClient } from '@scalekit-sdk/node'
   import 'dotenv/config'
   import { createInterface } from 'node:readline/promises'

   const scalekit = new ScalekitClient(
     process.env.SCALEKIT_ENVIRONMENT_URL,
     process.env.SCALEKIT_CLIENT_ID,
     process.env.SCALEKIT_CLIENT_SECRET,
   )
   const actions = scalekit.actions

   const connector = 'huntermcp'
   const identifier = 'user_123'

   // Generate an authorization link for the user
   const { link } = await actions.getAuthorizationLink({ connectionName: connector, identifier })
   console.log('Authorize Hunter MCP:', link)
   const rl = createInterface({ input: process.stdin, output: process.stdout })
   await rl.question('Press Enter after authorizing...')
   rl.close()

   // Make your first call
   const result = await actions.executeTool({
     connector,
     identifier,
     toolName: 'huntermcp_find_people',
     toolInput: {},
   })
   console.log(result)
   ```

   ```bash
   npx tsx quickstart.mts
   ```

   **Python** (`quickstart.py`)

   ```python
   import os
   from scalekit import ScalekitClient
   from dotenv import load_dotenv
   load_dotenv()

   scalekit_client = ScalekitClient(
       env_url=os.getenv("SCALEKIT_ENVIRONMENT_URL"),
       client_id=os.getenv("SCALEKIT_CLIENT_ID"),
       client_secret=os.getenv("SCALEKIT_CLIENT_SECRET"),
   )
   actions = scalekit_client.actions

   connection_name = "huntermcp"
   identifier = "user_123"

   # Generate an authorization link for the user
   link_response = actions.get_authorization_link(
       connection_name=connection_name,
       identifier=identifier,
   )
   print("Authorize Hunter MCP:", link_response.link)
   input("Press Enter after authorizing...")

   # Make your first call
   result = actions.execute_tool(
       tool_input={},
       tool_name="huntermcp_find_people",
       connection_name=connection_name,
       identifier=identifier,
   )
   print(result)
   ```

   ```bash
   python quickstart.py
   ```

   Each user signs in once. See [Authorize a user](/agentkit/tools/authorize/) for the full flow and statuses.

## Tools

Pass the exact name to `execute_tool`, with an input like each tool's example. To ask for a tool that's missing, use the [request form](https://scalekitsupport.portal.usepylon.com/forms/request-a-connector-tool).

**Node.js**

```typescript
const result = await actions.executeTool({
  toolName: 'huntermcp_email_count',
  toolInput: {
    domain: '<domain>',
  },
  connector: 'huntermcp',
  identifier: 'user_123',
})
```

**Python**

```python
result = actions.execute_tool(
    tool_name="huntermcp_email_count",
    tool_input={
        "domain": "<domain>",
    },
    connection_name="huntermcp",
    identifier="user_123",
)
```

`result.data` is the app's response as JSON, and `result.execution_id` (`executionId` in Node.js) is the ID of the call.

### `huntermcp_email_count`

Email Count · Read-only

Use this when the user wants the count of email addresses Hunter has indexed for a domain, optionally split by personal vs generic. Free to call. Do not use this as a substitute for Domain-Search; this returns a count only, not the email list itself.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `domain` (`string`, required): Domain name to count email addresses for
- `type` (`string`, optional): Type of email addresses to count. One of: `personal`, `generic`.

Example input: `{"domain":"<domain>"}`

### `huntermcp_find_companies`

Find Companies · Read-only

Use this when the user wants to search for companies that match natural-language criteria such as location, industry, size, type, or technologies. Returns up to 100 matching companies per page; use `limit` and `offset` to paginate. Plans without Discover paging get only the first page. The response includes `meta.permalink` — a link to the same query on hunter.io that the user can open to view all results with the inferred filters applied. Free to call.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `query` (`string`, required): Your search query
- `limit` (`integer`, optional): Maximum number of companies to return (default 100, max 100)
- `offset` (`integer`, optional): Number of companies to skip (default 0, max 10000). Plans without Discover paging get only the first page: any offset above 0 is rejected, so use `limit` for a smaller page.

Example input: `{"query":"fintech startups in France with 11-50 employees"}`

### `huntermcp_find_people`

Count People at Companies · Read-only

Count the people that can be extracted from companies — the bridge from Find-Companies to actual contacts; call it right after a Find-Companies result when the user says 'now get me the people at these companies'. For each matching company it reports how many email addresses Hunter's public index holds: `emails_count.personal` counts addresses tied to a named person (the people you can actually extract), `emails_count.generic` counts role-based addresses such as info@ or sales@, and `emails_count.total` combines both. `meta.total_emails` aggregates the same three counters across ALL companies matching the search — not just the current page — so use it to size a prospecting batch upfront. This tool returns counts only, from Hunter's public index: to reveal the actual email addresses, names, and positions at a company, call Domain-Search with that company's domain (uses credits). Provide either `query` (natural-language criteria, translated to filters exactly like Find-Companies) or `domains` (exact company domains, e.g. lifted from a Find-Companies result); when both are given, `query` takes precedence and `domains` is not sent. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `domains` (`array`, optional): Exact company domains to extract people counts for (e.g. from a Find-Companies result). Ignored when `query` is provided.
- `limit` (`integer`, optional): Maximum number of companies to return (default 100, max 100)
- `offset` (`integer`, optional): Number of companies to skip (default 0). Plans without Discover paging get only the first page: any offset above 0 is rejected, so use `limit` for a smaller page.
- `query` (`string`, optional): Natural-language company search criteria, translated to Discover filters exactly like Find-Companies. Example: 'fintech startups in France with 11-50 employees'

Example input: `{}`

### `huntermcp_find_similar_companies`

Find Similar Companies · Read-only

Find companies similar to a seed, ranked by `similarity_score` (0-100, higher is closer). Seed with exactly one of `company_list_id` (a static company list; List-Company-Lists shows each list's `type`) or `domain` (one company). `industry` and `size` are hard filters. `industry` must match a LinkedIn industry name exactly (case-insensitive), so reuse the `industry` values from earlier results. `country` is a ranking preference, not a filter: same-country companies rank higher, but other countries and companies with no known country can still appear, so check each result's `country` before saying all results are in one country. There is no paging. An empty `data` array means the domain is not in Hunter's index, the list has no companies, nothing similar was found, or the `industry`/`size` filters removed every result; say which is likely instead of saying nothing similar exists. A dynamic list returns an invalid_input error (field unsupported_list_type): use a static company list or a `domain` seed instead. This feature is not available on free plans. A feature_not_in_plan error means the user's plan does not include it: tell the user that and do not ask the user to reconnect. An unauthorized error is an authentication problem. To save results, call Save-Company for each chosen domain to get its company id. For one or two companies, then call Add-Company-To-List; for more, add all the ids in one Bulk-Copy-Companies call. A company that is already in the target list counts as saved (Add-Company-To-List returns `already_in_list: true`); continue with the next one. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `company_list_id` (`integer`, optional): ID of a static company list to seed from. Provide this or `domain`, not both.
- `country` (`string`, optional): ISO 3166 alpha-2 country code, e.g. FR. Ranks same-country companies higher; it does not filter.
- `domain` (`string`, optional): Domain or website URL of one company to seed from, e.g. stripe.com. Provide this or `company_list_id`, not both.
- `industry` (`array`, optional): Keep only these industries (exact LinkedIn industry names, case-insensitive)
- `limit` (`integer`, optional): Maximum number of companies to return (default 20, max 100)
- `size` (`array`, optional): Keep only these employee-count buckets

Example input: `{}`

### `huntermcp_get_account_details`

Get Account Details · Read-only

Use this when the user asks about their Hunter account, plan, team, or remaining credits. Returns plan name, team info, and per-product credit balances (search, verification, enrichment). Free to call.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs: none.

Example input: `{}`

### `huntermcp_get_briefing`

Get Briefing · Read-only

Get a short summary of what is new in the user's Hunter account. It answers open questions such as "What should I do in Hunter today?" or "What is new in Hunter?". It returns the number of new lead signal events (funding rounds, technology changes, new contacts and executives, job posts, hiring and headcount changes, and contacts who moved or changed position, in the team's lists), the number of new events from the team's Monitored Signals, the number of this week's opportunities (groups of people or companies like the ones the team saved), and the plan and remaining credits. Each item has a status: ok, unavailable (it could not be read now; try again later), or no_access (the plan does not include it, or a team setting turned it off; reason gives the error id). When count_exact is false, count is a lower bound. Each item's next_call names the tool (mcp_tool) and the arguments (params) that read its details, such as List-Signals or List-Opportunities. By default it counts the last 7 days; since counts only what happened after it, at most 30 days back. To check again later, pass meta.next_since as since. Free (no credits), and it marks nothing as seen.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `leads_since` (`string`, optional): Like since, for the lead signals item only: pass that item's last next_since.
- `monitored_since` (`string`, optional): Like since, for the Monitored Signals item only: pass that item's last next_since.
- `since` (`string`, optional): ISO 8601 date-time, not in the future, such as 2026-11-01T00:00:00Z. Count only what happened after it. Default: 7 days ago. A value older than 30 days reads as 30 days ago.

Example input: `{}`

### `huntermcp_get_company_list`

Get Company List · Read-only

Get a single company list by ID, including its `name`, `type` (static or dynamic), `filters` (for dynamic lists), `company_list_folder_id`, `created_at`, and `companies_count` (the number of companies currently in the list). Returns a not-found error if the list does not exist or belongs to another team. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`integer`, required): ID of the company list to retrieve

Example input: `{"id":1}`

### `huntermcp_get_connected_app`

Get Connected App · Read-only

Get a single connected app by ID, including its `attribute_mappings` — the list of `target_field` ↔ `source_field` pairs that map Hunter fields to the integration's fields. Each mapping may also include a `direction` (`export` to the integration, `import` into Hunter, or `both`) and a `conflict_policy` for imports (`crm_wins` overwrites the Hunter value, `fill_if_empty` only fills an empty one; null for export-only mappings). Also returns the `provider`, resolved `name`, `category`, `provider_email`, and `connected_at` / `updated_at` timestamps. Returns a not_found error if the app does not exist or belongs to a different team. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`integer`, required): ID of the connected app to retrieve

Example input: `{"id":1}`

### `huntermcp_get_custom_attribute`

Get Custom Attribute · Read-only

Get one custom attribute by its `id` from List-Custom-Attributes: its `label` and its `slug`. Lead records from Get-Lead and List-Leads show its value under the slug. To set a value on a lead, send `custom_attributes` keyed by this slug to Update-Lead. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`integer`, required): ID of the custom attribute to retrieve

Example input: `{"id":1}`

### `huntermcp_get_email_account`

Get Email Account · Read-only

Get the full configuration of one sending email account — the pre-check to run before suggesting changes to an account's settings. On top of the List-Email-Accounts fields (email, first/last name, `sending_status`, `daily_limit`, `provider`), it reports the `sender_name` shown on outgoing emails, whether it is the `default_account`, the email `signature` (sanitized HTML, null when none is stored), the owner's `profile_picture_url`, the team's `custom_tracking_domain` (null unless a verified custom tracking domain is configured), the default `sending_schedule` (days of week, HH:MM start/end times, and time zone) a new sequence sending from this account would run on, and the `warmup` state (`enabled`, `status`, `end_date`). Email account settings can NOT be changed through this API — inspection is read-only and settings writes are not yet available — so suggest changes for the user to apply in the hunter.io dashboard rather than offering to apply them. Returns a not-found error if the email account does not exist or is not one the user can manage. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `email_account_id` (`integer`, required): ID of the email account to inspect

Example input: `{"email_account_id":1}`

### `huntermcp_get_lead`

Get Lead · Read-only

Get the record of one lead by its lead `id`: contact and company fields (without `company_type`, which the API does not return), one of its lists (`leads_list`, only one even when the lead is in several), `sync_status`, `sending_status`, verification, and custom attribute values under each attribute `slug`. Get the `id` from Lead-Exists (by email) or List-Leads. Do not guess an id. Returns a not-found error if the lead does not exist or belongs to another team. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`integer`, required): ID of the lead to retrieve

Example input: `{"id":1}`

### `huntermcp_get_leads_list`

Get Leads List · Read-only

Get one leads list by its `id` from List-Leads-Lists: its `name`, `leads_count`, `type` (static or dynamic), `created_at`, and its first 20 leads. To read all the leads in the list, or to filter them, use List-Leads with `leads_list_id`. Only a static list can receive leads. Returns a not-found error if the list does not exist or belongs to another team. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`integer`, required): ID of the leads list to retrieve

Example input: `{"id":1}`

### `huntermcp_get_message_template`

Get Message Template · Read-only

Retrieve a single saved message template by ID — its full `subject`, `body`, and `message_format` — for example to preview it before reusing it in a sequence follow-up (via Create-Sequence-Follow-Up's `message_template_id`), or to copy and adapt its content. Returns a not-found error if the template does not exist or belongs to another team. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`integer`, required): ID of the message template to retrieve

Example input: `{"id":1}`

### `huntermcp_get_opportunity_results`

Get Opportunity Results · Read-only

List the people or companies of one opportunity from List-Opportunities, the closest match first; meta.type says which. People are masked: first name and last initial, position, department, seniority and company domain, with no email address. Each person has a reveal_handle: Reveal-Emails gets the email addresses of the people the user picks, at 1 credit per email. Companies come with their domain and the number of email addresses Hunter has for them (personal, generic, total). People the team already saved as leads are left out, so meta.results can be lower than the opportunity's results_count. Supports limit and offset. Free (no credits), and it marks nothing as seen.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `opportunity_id` (`integer`, required): The id of an opportunity from List-Opportunities
- `limit` (`integer`, optional): Maximum number of people or companies to return (default 20, max 100)
- `offset` (`integer`, optional): Number of people or companies to skip (default 0)

Example input: `{"opportunity_id":1}`

### `huntermcp_get_saved_search`

Get Saved Search · Read-only

Get one saved Discover search by ID — typically to rerun it. Read the stored `filters` payload and describe it back to the user as a natural-language `query` for Find-Companies or Find-People; this is an approximate rerun, because those tools accept only a `query` (or exact `domains`) and cannot re-apply the stored structured filters (locations, industries, funding, technologies, include/exclude lists) verbatim. Present the reformulated query and let the user confirm or refine before running it. Returns a not-found error if the saved search does not exist or belongs to another user. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `saved_search_id` (`integer`, required): ID of the saved search to fetch

Example input: `{"saved_search_id":1}`

### `huntermcp_get_sequence`

Get Sequence · Read-only

Get one sequence's full configuration. Returns the lifecycle `status` (draft, planned, active, paused, completed, preparing, error, or archived) plus the raw `started`/`paused`/`archived` booleans, the `schedule` (start_at date, daily sending window as HH:MM strings, days of week as integers 0=Monday..6=Sunday), the `settings` (open tracking, link tracking, unsubscribe link, the HTML and plain-text unsubscribe sentences — null when the account default applies, AI assistant, BCC recipient), the sender `email_account_ids`, a `follow_ups` step summary (unique_steps_count and step numbers — use List-Sequence-Follow-Ups for the step contents), `recipients_count`, the `owner`, and the `ab_testing` state (`enabled`, the Variant A/B split percentages, and any `winner_variant`; every field but `enabled` is null when no A/B test is on). Returns a not-found error if the sequence does not exist or belongs to another team. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence to fetch

Example input: `{"sequence_id":1}`

### `huntermcp_get_sequence_follow_up`

Get Sequence Follow-Up · Read-only

Get a single step of a sequence by its follow-up ID (from List-Sequence-Follow-Ups). Returns the step number (0 is the introduction email), `wait_days`, `subject`, `body`, `message_format` (text or html), `messages_sent`, and `variant` (null, or A/B for A/B-tested steps). Unlike List-Sequence-Follow-Ups, `subject` is returned raw: a step that inherits the previous step's subject shows an inheritance placeholder instead of the resolved text. Returns a not-found error if the sequence or the step does not exist or belongs to another team. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `follow_up_id` (`integer`, required): ID of the follow-up step to fetch (from List-Sequence-Follow-Ups)
- `sequence_id` (`integer`, required): ID of the sequence the step belongs to

Example input: `{"follow_up_id":1,"sequence_id":1}`

### `huntermcp_get_sequence_stats`

Get Sequence Stats · Read-only

Use this when the user wants to see how a sequence is performing. Reports recipient-based stats for a sequence plus a per-follow-up breakdown. The top-level engagement totals (`sent`, `delivered`, `opened`, `clicked`, `replied`, `unsubscribed_recipients`) count DISTINCT recipients — e.g. `replied` is the number of recipients who replied at least once, not the number of reply messages — and `delivered` is the recipients who received at least one non-bounced email (`recipients_count` is all recipients ever added to the sequence). `bounced` is message-based (a bounce is an SMTP-level event). The rate fields (`open_rate`, `click_rate`, `reply_rate`, `bounce_rate`, `unsubscribed_recipients_rate`) are floats in the 0.0-1.0 range — multiply by 100 to present them as percentages to the user — and are 0 when their denominator is 0; `open_rate`, `click_rate`, and `reply_rate` divide by distinct delivered recipients, while `bounce_rate` is message-based. The `follow_ups` array reports MESSAGE-based metrics per step (ordered by step, then variant; `variant` is null for the default follow-up or A/B for A/B-tested steps), so per-step counts do not sum to the recipient-based top-level totals. Returns a not-found error if the sequence does not exist or belongs to another team. Free to call.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence to fetch stats for

Example input: `{"sequence_id":1}`

### `huntermcp_get_usage`

Get Quota Usage · Read-only

Report the team's quota usage for the current billing period. Call this proactively BEFORE any large credit-consuming action (a Domain-Search loop over many companies, bulk email verification, bulk enrichment) and warn the user when the planned work would exceed the remaining quota. Paid results already report each call's charge and, when known, the credits left (`credits.remaining`), so do not call this between the calls of a batch. Returns `reset_date` (the day the quota resets, i.e. the end of the current billing period) and buckets, each with `used`, `available`, and `remaining`. How to read the buckets depends on the plan. When `requests.credits` is present, the plan has ONE shared credit pool: plan with `credits.remaining`. A search costs 1 credit, a verification 0.5 credit and an enrichment 0.2 credit (an enrichment still needs at least 1 credit left to start). `searches` and `verifications` are then not separate quotas but approximate views of the same pool (rounded down, with request-pack usage counted per type): do not add them up, and do not use them to tell what one call cost. When `credits` is absent, `searches` and `verifications` are separate quotas: Domain-Search, Email-Finder and enrichment use `searches`, Email-Verifier uses `verifications`. Reading usage consumes no credits and no requests. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs: none.

Example input: `{}`

### `huntermcp_lead_exists`

Check if Lead Exists · Read-only

Check if the team already has a lead with this email. Returns the lead `id` and the `leads_list_id` and `leads_list_name` of one of its lists, even when the lead is in several. All three are null when no lead has this email. A lead in no list also counts, for example one that automatic lead saving created. Use the `id` with Get-Lead, Update-Lead, Add-Tag-To-Lead, or Add-Sequence-Recipients. To save the person only if no lead exists, call Create-Lead-If-Missing directly. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `email` (`string`, required): Email address to check

Example input: `{"email":"<email>"}`

### `huntermcp_list_api_keys`

List API Keys · Read-only

List the Hunter API keys of the connected user. Keys are scoped to the requesting user, not the team — another team member's keys are never visible here. Key values are masked (only the last 4 characters are shown, e.g. `********abcd`); the full value of a key is visible only once, in the Create-API-Key response. Managing keys requires connecting with an API key: connections authorized via OAuth get an unauthorized error ("API keys can't be managed with an OAuth token. Use your API key instead.") — relay that message to the user instead of retrying. Supports `limit` and `offset` pagination. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of API keys to return (default 20, max 100)
- `offset` (`integer`, optional): Number of API keys to skip (default 0)

Example input: `{}`

### `huntermcp_list_company_list_folders`

List Company List Folders · Read-only

List the company-list folders in your Hunter account, ordered with the most recently created first. Folders group company lists; each folder reports its `name`, `color` (a hex color string), `company_lists_count` (how many lists are filed in it), and `created_at`. Folders are visible to the whole team. Supports `limit` and `offset` pagination. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of folders to return (default 20, max 100)
- `offset` (`integer`, optional): Number of folders to skip (default 0)

Example input: `{}`

### `huntermcp_list_company_lists`

List Company Lists · Read-only

List the company lists in your Hunter account, ordered with the most recently created first. Each list reports its `type` — `static` (a fixed set of companies you add manually) or `dynamic` (companies matched automatically by saved `filters`, which are included for dynamic lists) — along with its `name`, `company_list_folder_id` (the folder it lives in, or null), and `created_at`. Supports `limit` and `offset` pagination. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of company lists to return (default 20, max 100)
- `offset` (`integer`, optional): Number of company lists to skip (default 0)

Example input: `{}`

### `huntermcp_list_connected_apps`

List Connected Apps · Read-only

List the connected apps (third-party integrations such as HubSpot, Pipedrive, Google Sheets, or a custom SMTP/IMAP inbox) linked to your Hunter account. Each entry includes the `provider` and a resolved `name` (canonical names like Google Sheets come from Hunter's internal catalog; others fall back to the provider name), plus `category`, `provider_email`, and the `connected_at` / `updated_at` timestamps. Supports `limit` and `offset` pagination. Returns an empty list if no apps are connected. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of apps to return (default 20, max 100)
- `offset` (`integer`, optional): Number of apps to skip (default 0)

Example input: `{}`

### `huntermcp_list_custom_attributes`

List Custom Attributes · Read-only

List the team's custom attributes: the extra lead fields that the user defined. Each has an `id`, a `label`, and a `slug`. Get-Lead and List-Leads show each value under its slug. Create-Lead and Update-Lead set values with `custom_attributes`, keyed by slug. Call this before Create-Custom-Attribute, so that you reuse an attribute and do not make a duplicate. Custom attributes can be mapped to fields of a connected CRM. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs: none.

Example input: `{}`

### `huntermcp_list_email_account_sequences`

List Sequences for Email Account · Read-only

List the sequences that send from a given email account — the dependency check to run before pausing, reconfiguring, or disconnecting an account. Pairs with List-Email-Accounts (find the account and its id) and Get-Email-Account (inspect its configuration). Each entry reports the sequence `id`, `name`, `status` (draft, planned, active, paused, or archived), `recipients_allocated` (recipients assigned to this account within the sequence, canceled recipients excluded), and `emails_scheduled` (emails still waiting to be sent from this account), most recent first. Archived sequences are excluded by default; pass `include_archived: true` to include them. Supports `limit` and `offset` pagination. Returns a not-found error if the email account does not exist or is not one the user can manage. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `email_account_id` (`integer`, required): ID of the email account whose sequences should be listed
- `include_archived` (`boolean`, optional): Set to true to include archived sequences (excluded by default)
- `limit` (`integer`, optional): Maximum number of sequences to return (default 20, max 100)
- `offset` (`integer`, optional): Number of sequences to skip (default 0)

Example input: `{"email_account_id":1}`

### `huntermcp_list_email_accounts`

List Email Accounts · Read-only

List the email accounts (sending inboxes) connected to your Hunter account. This is the first step before creating or starting a sequence — every account reports a `sending_status` of `active` (ready to send), `paused` (disconnected, will not send until reconnected), or `warming` (in ramp-up or warmup, sending a reduced volume). Each entry also includes the sender's email, name, daily sending limit, and provider (such as gmail or outlook). Supports `limit` and `offset` pagination. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of email accounts to return (default 20, max 100)
- `offset` (`integer`, optional): Number of email accounts to skip (default 0)

Example input: `{}`

### `huntermcp_list_job_titles`

List Job Titles (Free) · Read-only

Use this when the user wants to see the job titles at one or more companies before spending credits, for example to choose who to contact at each company. Free: it returns no names or email addresses and uses no credits. For each domain (up to 25 in one call) it counts every person Hunter has: people_total, the people per department and per seniority ("unknown" when Hunter has no value), and the most common job titles, each with its department, seniority, number of people, and number of decision makers. people_in_other_titles counts the people whose titles did not fit in the list; to see them, raise limit or filter by department, seniority, or title_keywords. Then call Preview-People with job_titles set to the titles the user chose (at most 20), and the same department and seniority as here: it lists those people for free, and Reveal-Emails then gets the emails of the ones the user wants, at 1 credit per email. Domain-Search does not filter by job title. Leave the comma out of a title that has one ('Director Marketing' for 'Director, Marketing'), because a comma separates titles.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `domains` (`array`, required): Company domains to list, 1 to 25. For more companies, call again with the next 25.
- `department` (`string`, optional): Filter by department. Values: executive, it, finance, management, sales, legal, support, hr, marketing, communication, education, design, health, operations, product, research, consulting, administrative, procurement. Supports comma-separated multi-values (e.g. 'sales,marketing')
- `limit` (`integer`, optional): Job titles to list for each company, 1 to 1000. Default: 100 for up to 5 companies, then fewer per company (at least 10), so a call lists about 500 titles in all.
- `seniority` (`string`, optional): Filter by seniority level. Values: junior, senior, executive. Supports comma-separated multi-values (e.g. 'senior,executive')
- `title_keywords` (`string`, optional): Only count people whose job title contains one of these words. Comma-separated, at most 20, e.g. 'sales,marketing' or 'engineer'.

Example input: `{"domains":[]}`

### `huntermcp_list_lead_tags`

List Lead Tags · Read-only

List the lead tags in your Hunter account — call this BEFORE tagging leads or creating a new tag, so an existing tag can be reused instead of creating a near-duplicate. Tags label leads for organization and filtering and are shared with the whole team. Tags are ordered with the most recently created first; each tag reports its `name`, `color` (a hex color string), and `created_at`. Supports `limit` and `offset` pagination (default limit 25, max 100). Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of tags to return (default 25, max 100)
- `offset` (`integer`, optional): Number of tags to skip (default 0)

Example input: `{}`

### `huntermcp_list_leads`

List Leads · Read-only

List the team's saved leads that match filters such as replied, invalid email, country, saved date, tags, sequence and job title. A lead must match every filter that you send, and a lead matching any one value of a filter matches that filter. `*` means any value and `~` means no value. Get tag ids from List-Lead-Tags, sequence ids from List-Sequences, custom attribute slugs from List-Custom-Attributes, and list ids from List-Leads-Lists. `meta.count` is the number of leads that match. Each page holds at most 100 leads, newest first by default. Use `offset` to get the next page. Do not create a dynamic list only to read leads: call this tool. For one email address, Lead-Exists is faster. Each lead shows its custom attribute values under the attribute `slug`. Lead ids from this tool are the input of Add-Sequence-Recipients, Bulk-Move-Leads, and Push-Leads-To-CRM. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `company` (`string`, optional): A company name, or a list of names: a lead at any one of them matches
- `company_industry` (`string`, optional): Company industry
- `company_type` (`string`, optional): Company type, or * (any type) or ~ (no type). One of: `educational`, `educational institution`, `government agency`, `non profit`, `partnership`, `privately held`, `public company`, `self employed`, `self owned`, `sole proprietorship`, `*`, `~`.
- `confidence_score` (`string`, optional): A from/to range between 0 and 100, or * (any score) or ~ (no score)
- `country_code` (`string`, optional): The company's two-letter country code, e.g. US or FR, or * (any country) or ~ (no country). For where the person is, use location.
- `created_at` (`string`, optional): When the lead was saved: a preset or a from/to date range
- `custom_attributes` (`object`, optional): Keyed by a custom attribute slug from List-Custom-Attributes. The value is text to match, * (any value) or ~ (empty).
- `decision_maker` (`boolean`, optional): true keeps only the leads marked as decision makers
- `department` (`array`, optional): Departments, or * (any department) or ~ (no department). A lead in any one matches.
- `email` (`string`, optional): Email address
- `email_type` (`array`, optional): generic (a shared address such as contact@) or personal, or * (any type) or ~ (no type)
- `first_name` (`string`, optional): First name
- `last_activity_at` (`string`, optional): No description.
- `last_contacted_at` (`string`, optional): No description.
- `last_exported_at` (`string`, optional): No description.
- `last_name` (`string`, optional): Last name
- `leads_list_id` (`integer`, optional): ID of a static or dynamic list, from List-Leads-Lists. Only its leads match.
- `limit` (`integer`, optional): Maximum number of leads to return (max 100)
- `linkedin_url` (`string`, optional): LinkedIn profile URL
- `location` (`object`, optional): Where the person is: locations to include or exclude. For the company's country, use country_code.
- `offset` (`integer`, optional): Number of leads to skip
- `people_per_company` (`integer`, optional): At most this many leads from each company. The cap applies to the whole result, so page with offset as usual. meta.count counts the leads after the cap. It works on at most 10,000 matching leads: past that, narrow the filters.
- `phone_number` (`string`, optional): Phone number
- `position` (`array`, optional): Job titles. A lead matching any one matches.
- `position_keywords` (`array`, optional): Keywords to find in the job title, at most 64 words in all. Each keyword also matches its synonyms, e.g. cto matches Chief Technology Officer. The synonyms count toward the same 64 words: past them, a keyword matches only as typed. With position, a lead matching any title or keyword matches.
- `query` (`string`, optional): Text to search in first name, last name, company, website and email
- `sending_status` (`array`, optional): Sending statuses, or * (any status) or ~ (no status)
- `seniority_level` (`array`, optional): Seniority levels, or * (any level) or ~ (no level)
- `sequence_ids` (`string`, optional): Sequence IDs from List-Sequences, or * (in any sequence) or ~ (in no sequence). Add ~ to the IDs to also keep leads in no sequence.
- `sort` (`string`, optional): newest (the default), oldest, or relevance: the best position or position_keywords matches first. One of: `newest`, `oldest`, `relevance`.
- `source` (`string`, optional): Source of the lead
- `sync_status` (`string`, optional): CRM sync status, or * (any status) or ~ (no status). One of: `pending`, `error`, `success`, `*`, `~`.
- `tags` (`array`, optional): Tag IDs from List-Lead-Tags
- `twitter` (`string`, optional): Twitter/X handle
- `updated_at` (`string`, optional): No description.
- `user_id` (`array`, optional): IDs of the team members who saved the leads. A lead's created_by holds this ID.
- `verification_date` (`string`, optional): No description.
- `verification_status` (`array`, optional): Email verification statuses, or * (any status) or ~ (no status)
- `website` (`string`, optional): Company website

Example input: `{}`

### `huntermcp_list_leads_list_folders`

List Leads List Folders · Read-only

List the leads-list folders in your Hunter account, ordered with the most recently created first. Folders group leads lists and are visible to the whole team — surface them when the account has many lists, since grouping lists by folder is how users keep a large workspace navigable. Each folder reports its `name`, `color` (a hex color string), `leads_lists_count` (how many lists are filed in it), and `created_at`. Supports `limit` and `offset` pagination (default 20, max 100). Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of folders to return (default 20, max 100)
- `offset` (`integer`, optional): Number of folders to skip (default 0)

Example input: `{}`

### `huntermcp_list_leads_lists`

List Leads Lists · Read-only

List the team's leads lists, newest first, with the `id`, `name`, `leads_count`, and `type` (static or dynamic) of each. Use it to find the `leads_list_id` for a list name that the user gives. Create-Lead, List-Leads, Bulk-Move-Leads, and Push-Leads-To-CRM take this id. Only a static list can receive leads. Use `limit` (at most 100) and `offset` to page. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of lists to return
- `offset` (`integer`, optional): Number of lists to skip

Example input: `{}`

### `huntermcp_list_message_templates`

List Message Templates · Read-only

List the saved message templates in the user's Hunter account. When the user is drafting a sequence step (an introduction message or a follow-up), list the templates first and ask whether they want to use a saved template as the starting point. Each template carries a reusable `subject`, `body`, and `message_format` (text or html). To build a follow-up step from one, pass the chosen template's id as `message_template_id` to Create-Sequence-Follow-Up: the step's subject, body, and format are pre-filled from the template (any subject/body passed explicitly wins over the template's). Supports `limit` and `offset` pagination (default 25 per page, max 100); templates are ordered newest first. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of templates to return (default 25, max 100)
- `offset` (`integer`, optional): Number of templates to skip (default 0)

Example input: `{}`

### `huntermcp_list_monitored_signals`

List Monitored Signals · Read-only

List the team's Monitored Signals, newest first, as on the Signals page: funding signals, which follow the funding rounds that match their filters (country, company size, industry, series, amount and date window); job openings signals, which follow the job posts that match their filters (company, job title and description words, job location, seniority, contract type, department and date window); and company follows, which follow the events of one company. Each signal gives its id, name, type (funding_rounds, job_openings or company_follow), status (active or paused), filters, and the event types its feed can return. Read a signal's events with List-Signals, source=monitored and its id as signal_id; a paused signal can be read only that way. Supports limit and offset. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of Monitored Signals to return (default 20, max 100)
- `offset` (`integer`, optional): Number of Monitored Signals to skip (default 0)

Example input: `{}`

### `huntermcp_list_opportunities`

List Opportunities · Read-only

List this week's opportunities: groups of people or companies that look like the ones the team saved, which Hunter shows as "New leads for you" on the dashboard. Each opportunity gives its type (people or companies), the source it is based on (the team's saved leads, the people who replied to its sequences, its saved companies, or recently funded companies like them), a description, results_count, and a score (higher is a closer match). The best come first: people before companies, then by score. Open one with Get-Opportunity-Results and its id. For about a day after the weekly refresh, the list can hold opportunities from both weeks. Supports limit and offset. Free (no credits), and it marks nothing as seen.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of opportunities to return (default 20, max 100)
- `offset` (`integer`, optional): Number of opportunities to skip (default 0)

Example input: `{}`

### `huntermcp_list_saved_searches`

List Saved Searches · Read-only

List the Discover searches saved in your Hunter account — a good opening move at the start of a prospecting conversation ('want to rerun one of your saved searches?'). Each saved search carries its `id`, `name`, the stored `filters` payload, and timestamps, ordered newest first. To rerun one, read its `filters` and `name` and describe them to the user as a natural-language `query` for Find-Companies or Find-People — that reformulation is approximate, since these tools take only a `query` (or exact `domains`) and cannot re-apply the stored structured filters (locations, industries, funding, technologies, include/exclude lists) verbatim. Supports `limit` and `offset` pagination (`meta.total` is the full count). Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of saved searches to return (default 25, max 100)
- `offset` (`integer`, optional): Number of saved searches to skip (default 0)

Example input: `{}`

### `huntermcp_list_sequence_follow_ups`

List Sequence Follow-Ups · Read-only

List the follow-up steps of a sequence, ordered by step. Step 0 is the introduction message and steps 1 and up are follow-ups; each step reports its `wait_days` (days to wait before sending), `subject` (the email subject; an inherited subject is resolved to the prior step's subject, so it is rarely empty), `body`, `message_format` (text or html), `messages_sent`, and `variant` (null, or A/B for A/B-tested steps). Supports `limit` and `offset` pagination. Returns 404 if the sequence does not exist or belongs to another team. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence to fetch follow-up steps for
- `limit` (`integer`, optional): Maximum number of follow-up steps to return (default 20, max 100)
- `offset` (`integer`, optional): Number of follow-up steps to skip (default 0)

Example input: `{"sequence_id":1}`

### `huntermcp_list_sequence_recipients`

List Sequence Recipients · Read-only

List the recipients of one sequence, with at most 100 for each page. Get `sequence_id` from List-Sequences. Each recipient has `email`, name, company, `sending_status`, and `lead_id`. Use it to see who the sequence emails. For open, click, and reply totals, use Get-Sequence-Stats. To add or stop recipients, use Add-Sequence-Recipients or Remove-Sequence-Recipients. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence
- `limit` (`integer`, optional): Maximum number of recipients to return
- `offset` (`integer`, optional): Number of recipients to skip

Example input: `{"sequence_id":1}`

### `huntermcp_list_sequences`

List Sequences · Read-only

List the sequences in your Hunter account with name, lifecycle booleans (started/archived/paused), recipient count, and owner, most recent first. Optionally filter by `started` and/or `archived` state and paginate with `limit`/`offset`. New sequences can be created with Create-Sequence and inspected in full with Get-Sequence. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `archived` (`boolean`, optional): Filter by archived state: true returns only archived sequences, false only non-archived ones
- `limit` (`integer`, optional): Maximum number of sequences to return (default 20, max 100)
- `offset` (`integer`, optional): Number of sequences to skip (default 0)
- `started` (`boolean`, optional): Filter by started state: true returns only sequences that have been started, false only drafts that were never started

Example input: `{}`

### `huntermcp_list_signals`

List Signals · Read-only

List signal events, newest first, from the last 30 days: the events the Hunter dashboard and Signals page show. With source=leads, the events are on companies and contacts in the team's lists: funding rounds, technologies added or removed, new contacts, new executives, job posts, hiring volume and hiring spikes, headcount growth, and contacts who changed location or position. With source=monitored, the events come from the team's Monitored Signals (funding rounds, job openings and company follows): one signal with signal_id (from List-Monitored-Signals), or else the team's active signals together. Each event gives the company domain, the event type and date, and its details: round type, amount in USD and date for a funding round; the technology for a technology change; the title for a new executive; the job title, location, country, employer link and whether it is still open for a job post; the open roles and role groups for hiring; the growth percent and headcounts for headcount growth; for a contact change, the previous and new value (from, to) and lead_id, the team's lead for the contact (read it with Get-Lead). Events carry no email address and no person name; next_call names Domain-Search, which finds people at the company and uses credits. hiring_volume, hiring_spike and company_growth are in early access: they come only for accounts that have them, and meta.types_skipped names the requested ones that the account does not have. Filter with types and since. Pages by cursor: pass meta.next_cursor as cursor, with the same source, signal_id, since and types. A page holds limit events unless it reached its read budget; then it can hold fewer and still have a next cursor. With source=monitored and no signal_id, meta.signals_skipped lists the active signals that the merged feed did not read; read each one with its signal_id. To check again later, pass meta.next_since as since. Free (no credits), and it marks nothing as seen.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `source` (`string`, required): leads: events on companies in the team's lists. monitored: events from the team's Monitored Signals. One of: `leads`, `monitored`.
- `cursor` (`string`, optional): meta.next_cursor of the previous page. Valid only with the same source, signal_id, since and types.
- `limit` (`integer`, optional): Maximum number of events to return (default 20, max 100)
- `signal_id` (`integer`, optional): With source=monitored only: the id of one Monitored Signal from List-Monitored-Signals, active or paused. Without it, the team's active signals are read together.
- `since` (`string`, optional): ISO 8601 date-time, not in the future. Return only events that occurred after it. A value older than 30 days reads as 30 days ago.
- `types` (`array`, optional): Event types to return. Default: all.

Example input: `{"source":"leads"}`

### `huntermcp_list_webhooks`

List Webhooks · Read-only

List the webhooks configured on the Hunter account — useful for integration debugging: is my webhook configured, which event does it listen to, and where does it deliver? Each webhook reports its `id`, `target_url` (the endpoint Hunter POSTs the event payload to), and `event` — one of lead.created, message.sent, message.read, message.clicked, message.replied, export.completed, import.completed, sequence.paused, or sequence.resumed. Results are ordered newest first and support `limit` and `offset` pagination (`meta.total` carries the overall count). Returns an empty list when no webhooks are configured. Free (no credits).

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `limit` (`integer`, optional): Maximum number of webhooks to return (default 20, max 100)
- `offset` (`integer`, optional): Number of webhooks to skip (default 0)

Example input: `{}`

### `huntermcp_plan_prospecting_flow`

Plan Prospecting Flow · Read-only

Use this when the user gives a natural-language prospecting brief (e.g. "Find 20 marketing leads at fintech companies in Berlin") and wants a step-by-step plan for finding companies and their contacts. Returns the plan plus a first action so the model can execute the chain end to end. By default it gathers contacts for review and returns them as a table; it only saves to the user's Hunter leads when the user explicitly asks (Hunter's automatic lead saving can still save what Email-Verifier checks — see that tool). Free to call; sub-tools charge their own credits.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `query` (`string`, required): Natural-language prospecting brief, e.g. 'Find 20 marketing leads at fintech companies in Berlin'.

Example input: `{"query":"Find 20 marketing leads at fintech companies in Berlin"}`

### `huntermcp_preview_people`

Preview People (Free) · Read-only

Use this when the user wants to see which people at one or more companies match the job titles they chose, before paying for any email address. Free: it uses no credits and returns no email addresses. For each domain (up to 25 in one call) it returns people_total and the matching people, masked: first name and last initial, position, department, seniority, decision_maker, verification status, and a reveal_handle. Filter with job_titles (the titles from List-Job-Titles), department, and seniority. Show the people and let the user choose who to contact; then call Reveal-Emails with their reveal_handle values, which costs 1 credit per email. A handle is valid for this team until the end of the billing period. When a company's next_search_after is not null, it has more people: call again with only that domain and search_after.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `domains` (`array`, required): Company domains to preview, 1 to 25. For more companies, call again with the next 25.
- `department` (`string`, optional): Filter by department. Values: executive, it, finance, management, sales, legal, support, hr, marketing, communication, education, design, health, operations, product, research, consulting, administrative, procurement. Supports comma-separated multi-values (e.g. 'sales,marketing')
- `job_titles` (`string`, optional): Only return people whose job title matches one of these titles. Comma-separated, at most 20, e.g. 'Head of Communications,Safeguarding Lead'. Every word of a title must match a word of the person's title, in any order, and common executive titles match each other (CTO matches Chief Technology Officer). Use the titles List-Job-Titles returned. A comma separates titles, so leave the comma out of a title that has one ('Director Marketing' for 'Director, Marketing').
- `limit` (`integer`, optional): People to return for each company, 1 to 100. Default: 25.
- `search_after` (`string`, optional): A company's next_search_after from the previous call, to get its next people. Only with exactly that one domain and the same filters.
- `seniority` (`string`, optional): Filter by seniority level. Values: junior, senior, executive. Supports comma-separated multi-values (e.g. 'senior,executive')

Example input: `{"domains":[]}`

### `huntermcp_add_company_to_list`

Add Company to List · Write

Add a company to a static company list (save the company to the list). Identify the list by `company_list_id` and the company by `company_id` (a company already saved in your account). Both the list and the company must belong to your team, and the list must be static — dynamic lists collect companies automatically by their filters and cannot have companies added manually. Returns the company (`id`, `domain`, `created_at`, `already_in_list`). `already_in_list: false` means the company was added; `already_in_list: true` means it was already in the list and nothing changed. Treat both as saved, not as an error. An invalid_input error saying the company "has already been taken" also means it is already in the list; treat it as saved too. Returns a not-found error if the list or company does not exist, belongs to another team, or the list is dynamic, or a forbidden error if you are not allowed to modify this list. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `company_id` (`integer`, required): ID of the company (already saved in your account) to add
- `company_list_id` (`integer`, required): ID of the static company list to add the company to

Example input: `{"company_id":1,"company_list_id":1}`

### `huntermcp_add_tag_to_lead`

Add Tag to Lead · Write

Add a tag to a lead, identified by `lead_id`. Provide either `tag_id` (an existing tag) or `tag_name`; with `tag_name`, a tag that does not exist yet is created automatically with a random color. Call List-Lead-Tags first and reuse an existing tag (by `tag_id` or its exact name) before introducing a near-duplicate name. The call is idempotent: re-adding a tag the lead already has still succeeds. Returns the lead's full current tag set (`tags`, each with `id`, `name`, and `color`). Returns an invalid_input error (id tag_required) if neither `tag_id` nor `tag_name` is given, a not-found error if the lead or the tag does not exist or belongs to another team, or an invalid_input error (id validation_failed) if the new tag name is invalid. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `lead_id` (`integer`, required): ID of the lead to tag
- `tag_id` (`integer`, optional): ID of an existing tag to attach (preferred when the tag already exists)
- `tag_name` (`string`, optional): Name of the tag to attach. Reuses the team's existing tag with that exact name, or creates it with a random color when missing. Ignored when tag_id is given.

Example input: `{"lead_id":1}`

### `huntermcp_bulk_domain_search`

Bulk Domain Search · Write

Use this when the user wants the contacts published at several companies, for example a list of companies they gave or picked from Find-Companies. One call searches up to 50 domains with the same filters (type, seniority, department, required field) and returns each company's emails with names, positions, confidence scores, and verification status. The searches run on Hunter's side, so the user approves this one call instead of one Domain-Search call per company. Costs like Domain-Search: 1 search per 10 emails returned for each company (rounded up), charged only when emails are found; on plans with a shared credit pool a search is 1 credit (see Get-Usage). So a batch costs at most the number of companies times limit/10 rounded up (10 companies at the default limit of 10: at most 10 searches). Before a call with more than one domain, tell the user how many companies it searches and the most it can cost, and get one approval for the whole batch; then set confirmed_credit_use. For more than 50 companies, split the list into calls of 50; that one approval covers every call of the batch. If a call stops early (out of credits, rate limited), the companies it did not search have status not_searched. It does not filter by job title: for that, use List-Job-Titles and Preview-People (both free), then Reveal-Emails. Do not use it for personal/webmail domains (gmail.com, yahoo.com, etc.). Saves nothing to the user's leads.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `domains` (`array`, required): Company domains to search, 1 to 50. For more companies, call again with the next 50.
- `confirmed_credit_use` (`boolean`, optional): Set true after the user approved the credit cost of this batch. Required with more than one domain; without it the server asks for the approval and searches nothing.
- `department` (`string`, optional): Filter by department. Values: executive, it, finance, management, sales, legal, support, hr, marketing, communication, education, design, health, operations, product, research, consulting, administrative, procurement. Supports comma-separated multi-values (e.g. 'sales,marketing')
- `job_titles` (`string`, optional): Not accepted: this tool does not filter by job title, and a call with job_titles is refused before anything is charged. To pick people by job title, use Preview-People (free, no email addresses), then Reveal-Emails (1 credit per email).
- `limit` (`integer`, optional): Maximum number of email addresses to return for each company, 1 to 100. Default: 10, which costs at most 1 search per company.
- `required_field` (`string`, optional): Only return results where this field has a value. One of: `full_name`, `position`, `phone_number`.
- `seniority` (`string`, optional): Filter by seniority level. Values: junior, senior, executive. Supports comma-separated multi-values (e.g. 'senior,executive')
- `type` (`string`, optional): Type of email addresses to return. One of: `personal`, `generic`.

Example input: `{"domains":[]}`

### `huntermcp_combined_enrichment`

Combined Enrichment · Write

Use this when the user provides an email address or LinkedIn handle and wants both the person's profile and their company's profile in a single response. Costs 0.2 credit on plans with a shared credit pool (but needs at least 1 credit left to start), or 1 search on other plans, charged only when data is found.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `email` (`string`, optional): Email address of the person to enrich
- `linkedin_handle` (`string`, optional): LinkedIn handle of the person to enrich (e.g. john-doe-123)

Example input: `{}`

### `huntermcp_company_enrichment`

Company Enrichment · Write

Use this when the user wants to look up a company by domain and see its industry, size, location, technologies, funding rounds, and social profiles. Does not return personal data (PII). Costs 0.2 credit on plans with a shared credit pool (but needs at least 1 credit left to start), or 1 search on other plans, charged only when data is found.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `domain` (`string`, required): Domain name of the company to enrich

Example input: `{"domain":"<domain>"}`

### `huntermcp_create_company_list`

Create Company List · Write

Create a new company list. Provide a `name`. By default the list is `static` (you add companies to it yourself). To create a `dynamic` list that automatically collects companies matching saved criteria, set `type` to `dynamic` and supply `filters` (a set of key/value criteria). Optionally place the list inside a folder with `company_list_folder_id`. Returns an invalid_input error (id validation_failed) if the name is missing, is already used by another list in your team, or a dynamic list is created without valid filters. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `name` (`string`, required): Name of the new company list
- `company_list_folder_id` (`integer`, optional): ID of the folder to place the list in (optional)
- `filters` (`object`, optional): Saved matching criteria for a dynamic list, keyed by Leads > Companies filter name. Each value is a string, number or boolean, or a list of them, never an object: there is no include/exclude form. Required when type is 'dynamic'.
- `type` (`string`, optional): List type: 'static' (manual, the default) or 'dynamic' (auto-matched by filters). One of: `static`, `dynamic`.

Example input: `{"name":"<name>"}`

### `huntermcp_create_company_list_folder`

Create Company List Folder · Write

Create a new company-list folder. Provide a `name` and a `color`. `color` must be one of the allowed Hunter folder colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B. Returns an invalid_input error (id validation_failed) if the name is missing or already used by another folder in your team, or if the color is missing or not one of the allowed values. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `color` (`string`, required): Folder color. One of the allowed Hunter folder colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B. One of: `374151`, `3489F9`, `10B981`, `F5BA0B`, `EF4444`, `7C3AED`, `F97316`, `E5E7EB`, `B4D9F7`, `BAE6B0`, `FDE68A`, `FBD0D0`, `D4C4F8`, `FFD79B`.
- `name` (`string`, required): Name of the new folder

Example input: `{"color":"374151","name":"<name>"}`

### `huntermcp_create_custom_attribute`

Create Custom Attribute · Write

Define a new custom attribute for leads, with a `label` such as "Instagram handle". Hunter makes a `slug` from the label. This defines the field only: it sets no value on any lead. Set values with the `custom_attributes` field of Create-Lead or Update-Lead, keyed by this slug. Call List-Custom-Attributes first, and reuse an attribute with the same meaning. A label that matches a built-in lead field is rejected. A team can have at most 64 custom attributes. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `label` (`string`, required): Label for the new custom attribute

Example input: `{"label":"<label>"}`

### `huntermcp_create_lead`

Create Lead · Write

Save one person as a lead in the team's Hunter account, identified by `email`. If a lead with this email already exists, the call succeeds and returns that lead, but does not apply the supplied values, including `custom_attributes`. It adds the lead to `leads_list_id` when you pass one, and the lead stays in its other lists. If you send no `leads_list_id` and the lead, new or existing, is in no list, Hunter adds it to the user's current list, if the user has one. Send the id of a static list from List-Leads-Lists: a dynamic list cannot hold leads that you add. Send Create-Lead calls one at a time. If a call fails with a `rate_limited` error, another lead is being created for the team: wait `retry_after_seconds` and send the same call again. To change the values of an existing lead, use Update-Lead or Create-Or-Update-Lead. To get an `alreadyExisted` flag and leave listed leads unchanged, use Create-Lead-If-Missing. A saved lead can then go to a sequence (Add-Sequence-Recipients) or to the user's CRM (Push-Leads-To-CRM). Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `email` (`string`, required): Email address of the lead (required)
- `company` (`string`, optional): Company name
- `company_industry` (`string`, optional): Industry of the company
- `company_size` (`string`, optional): Size range of the company (e.g. 1-10, 11-50)
- `company_type` (`string`, optional): Type of company. Leave it out if you are not sure. An empty string clears it on an existing lead. One of: `educational`, `educational institution`, `government agency`, `non profit`, `partnership`, `privately held`, `public company`, `self employed`, `self owned`, `sole proprietorship`, ``.
- `country_code` (`string`, optional): Two-letter country code (e.g. US, FR)
- `custom_attributes` (`object`, optional): Custom attribute values, keyed by slug (not by label). Get each slug from List-Custom-Attributes. The attribute must already exist: create it with Create-Custom-Attribute first. An unknown slug fails the call.
- `first_name` (`string`, optional): First name of the lead
- `last_name` (`string`, optional): Last name of the lead
- `leads_list_id` (`integer`, optional): ID of the leads list to add the lead to
- `linkedin_url` (`string`, optional): LinkedIn profile URL
- `notes` (`string`, optional): Notes about the lead
- `phone_number` (`string`, optional): Phone number
- `position` (`string`, optional): Job title or position
- `source` (`string`, optional): Source of the lead
- `twitter` (`string`, optional): Twitter/X handle
- `website` (`string`, optional): Company website URL

Example input: `{"email":"<email>"}`

### `huntermcp_create_lead_if_missing`

Create Lead if Missing · Write

Use this when the user wants to save a verified contact as a new lead without modifying any existing lead. If a lead with the email already exists, returns it with `alreadyExisted: true` and never applies the supplied values to it; otherwise creates the lead with the supplied fields. Custom attribute values apply only when this tool creates the lead. An existing lead in no list (for example, one the team's automatic lead saving stored when Email-Verifier ran) is added to `leads_list_id` when you pass one, and Hunter can then refresh its fields from Hunter's own data, as for any lead saved in Hunter. A lead already in a list is never moved, added to another list, or changed. Use Create-Or-Update-Lead instead when the user explicitly wants to overwrite an existing lead. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `email` (`string`, required): Email address of the contact to save
- `company` (`string`, optional): Company name
- `company_industry` (`string`, optional): Industry of the company
- `company_size` (`string`, optional): Size range of the company (e.g. 1-10, 11-50)
- `company_type` (`string`, optional): Type of company. Leave it out if you are not sure. An empty string clears it on an existing lead. One of: `educational`, `educational institution`, `government agency`, `non profit`, `partnership`, `privately held`, `public company`, `self employed`, `self owned`, `sole proprietorship`, ``.
- `confirmed_credit_use` (`boolean`, optional): Loop carry: forwarded from Domain-Search bulk credit-consent guard.
- `confirmed_save_use` (`boolean`, optional): Loop carry: forwarded from Domain-Search save-batch credit-consent guard.
- `country_code` (`string`, optional): Two-letter country code (e.g. US, FR)
- `custom_attributes` (`object`, optional): Custom attribute values, keyed by slug (not by label). Get each slug from List-Custom-Attributes. The attribute must already exist: create it with Create-Custom-Attribute first. An unknown slug fails the call.
- `department` (`string`, optional): Loop carry: forwarded from Domain-Search.
- `first_name` (`string`, optional): First name of the lead
- `job_titles` (`string`, optional): Loop carry: forwarded from Domain-Search, which refuses it.
- `last_name` (`string`, optional): Last name of the lead
- `leads_list_id` (`integer`, optional): ID of the leads list to add the lead to
- `limit` (`number`, optional): Loop carry: forwarded from Domain-Search.
- `linkedin_url` (`string`, optional): LinkedIn profile URL
- `notes` (`string`, optional): Notes about the lead
- `pending_companies` (`array`, optional): Multi-company loop carry (capped at 50 domains). When the user has picked 2+ companies, set this on the first Domain-Search call to the array of REMAINING picks (everything except the first picked domain that goes in the 'domain' arg). The response chain threads through Domain-Search per company and auto-continues to the next pending domain without between-company confirmation gates; when saving it also routes Email-Verifier → Create-Lead-If-Missing. Downstream tools receive this via suggestedArgs and shrink it as the chain advances; pass it through verbatim from the prior tool's chained nextAction. Leave UNDEFINED for single-call usage and for prospecting flows where the user picked exactly ONE company (single-company mode preserves the confirmation gate on Email-Verifier).
- `phone_number` (`string`, optional): Phone number
- `position` (`string`, optional): Job title or position
- `required_field` (`string`, optional): Loop carry: forwarded from Domain-Search. One of: `full_name`, `position`, `phone_number`.
- `save_leads` (`boolean`, optional): Loop carry: forwarded from Domain-Search research-vs-save mode.
- `seniority` (`string`, optional): Loop carry: forwarded from Domain-Search.
- `source` (`string`, optional): Source of the lead
- `twitter` (`string`, optional): Twitter/X handle
- `type` (`string`, optional): Loop carry: forwarded from Domain-Search. One of: `personal`, `generic`.
- `website` (`string`, optional): Company website URL

Example input: `{"email":"<email>"}`

### `huntermcp_create_lead_tag`

Create Lead Tag · Write

Create a new lead tag. Before creating one, call List-Lead-Tags and prefer reusing an existing tag with the same meaning (e.g. "Customers" vs "customer") instead of creating a near-duplicate — Add-Tag-To-Lead can also create a tag on the fly by name while tagging. Provide a `name`; `color` is optional and must be one of the allowed Hunter tag colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B — when omitted, Hunter picks a random color from that palette. Tags are shared with the whole team. Returns an invalid_input error (id validation_failed) if the name is missing, longer than 255 characters, or already used by another lead tag in your team. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `name` (`string`, required): Name of the new lead tag (unique within the team)
- `color` (`string`, optional): Tag color. One of the allowed Hunter tag colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B. Omit to let Hunter pick a random color from that palette. One of: `374151`, `3489F9`, `10B981`, `F5BA0B`, `EF4444`, `7C3AED`, `F97316`, `E5E7EB`, `B4D9F7`, `BAE6B0`, `FDE68A`, `FBD0D0`, `D4C4F8`, `FFD79B`.

Example input: `{"name":"<name>"}`

### `huntermcp_create_leads_list`

Create Leads List · Write

Create a new leads list with a `name`, and get its `id`. The user cannot have two lists with the same name. By default the list is static: it keeps the leads you add to it. Save leads into a static list with Create-Lead and `leads_list_id`, or move existing leads into it with Bulk-Move-Leads. You can push the list to the user's CRM with Push-Leads-To-CRM. To create a dynamic list (a segment), set `type` to `dynamic` and give `filters`: the list then updates itself. It holds every lead that matches the filters, also leads saved later, and a lead that stops matching leaves it. Leads can't be added to a dynamic list or removed from it by hand. To read leads, use List-Leads: create a dynamic list only when the user asks for a list that updates itself. A filter with the same name as a List-Leads filter works the same way. `leads_count` in the result is the number of leads that match. Optionally place the list inside a folder with `leads_list_folder_id` (see List-Leads-List-Folders). Returns an invalid_input error if the folder does not belong to your team, if a filter is not supported, if `filters` is empty, or if `leads_list_id` is a dynamic list. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `name` (`string`, required): Name of the new leads list
- `filters` (`object`, optional): Required when type is 'dynamic': the leads the list holds
- `leads_list_folder_id` (`integer`, optional): ID of the folder to place the list in (optional)
- `type` (`string`, optional): 'static' (the default): you add leads yourself. 'dynamic': the list holds the leads matching filters. One of: `static`, `dynamic`.

Example input: `{"name":"<name>"}`

### `huntermcp_create_leads_list_folder`

Create Leads List Folder · Write

Create a new leads-list folder to group leads lists — worth suggesting once the account has many lists. Provide a `name`; `color` is optional and must be one of the allowed Hunter folder colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B — when omitted, Hunter picks a random color from that palette. A team can have at most 100 folders. Returns an invalid_input error (id validation_failed) if the name is missing or already used by another folder in your team, if the color is not one of the allowed values, or if the folder limit is reached. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `name` (`string`, required): Name of the new folder (unique within the team)
- `color` (`string`, optional): Folder color. One of the allowed Hunter folder colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B. Omit to let Hunter pick a random color from that palette. One of: `374151`, `3489F9`, `10B981`, `F5BA0B`, `EF4444`, `7C3AED`, `F97316`, `E5E7EB`, `B4D9F7`, `BAE6B0`, `FDE68A`, `FBD0D0`, `D4C4F8`, `FFD79B`.

Example input: `{"name":"<name>"}`

### `huntermcp_create_message_template`

Create Message Template · Write

Create a reusable message template — including saving a message draft written in chat (e.g. a sequence introduction or follow-up the user liked) as a new template for later reuse. `name` and `body` are required; `subject` is optional and `message_format` defaults to html. The saved template can then pre-fill sequence steps via Create-Sequence-Follow-Up's `message_template_id`. Personalization placeholders like {{first_name}} are supported in the subject and body, and a fallback goes after a colon: {{first_name:"there"}}. The colon is the only separator that works — everything before it is read as the attribute name, so a piped form like {{first_name|fallback:"there"}} silently renders as an empty string. Placeholders carrying the literal fallback markers FALLBACK or DEFAULT are rejected with a validation error. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `body` (`string`, required): Message body (max 50,000 characters); HTML unless message_format is text. Personalization placeholders take a fallback after a colon, as in {{first_name:"there"}}
- `name` (`string`, required): Name of the template, shown in the Hunter template picker
- `message_format` (`string`, optional): Message format of the body; defaults to html when omitted. One of: `html`, `text`.
- `subject` (`string`, optional): Email subject line (max 250 characters). May include personalization placeholders; give each one a fallback after a colon, as in {{first_name:"there"}}

Example input: `{"body":"<body>","name":"<name>"}`

### `huntermcp_create_saved_search`

Create Saved Search · Write

Create a saved Discover search: store the current filter set under a name so the search can be rerun later. Stores the `filters` payload verbatim and returns the created saved search; it also appears on the hunter.io Discover page. Names must be unique among your saved Discover searches — a duplicate name returns a validation error. There is no update endpoint: to change a saved search, delete it with Delete-Saved-Search and recreate it. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `name` (`string`, required): Name for the saved search (must be unique among your saved Discover searches)
- `filters` (`object`, optional): Discover filter payload to store verbatim — the same shape returned in `filters` by Get-Saved-Search or List-Saved-Searches. Values should be strings, numbers, booleans, or arrays of those. Omit to save an empty filter set.

Example input: `{"name":"<name>"}`

### `huntermcp_create_sequence`

Create Sequence · Write

Create a new outreach sequence. Only `name` is required; the sequence is created as a DRAFT that sends nothing until it is started. Lifecycle: draft → started (Start-Sequence) → paused/archived — the schedule and sender fields lock once started, so configure them while drafting. Optional fields: sender `email_account_ids` (connected email accounts — see List-Email-Accounts; unknown ids are rejected with unknown_email_account_ids), `schedule_days` (0=Monday..6=Sunday; a Monday-Friday schedule is [0, 1, 2, 3, 4], not [1, 2, 3, 4, 5]), the daily sending window `schedule_time_start`/`schedule_time_end` (seconds since midnight, start before end), a `start_at` date (YYYY-MM-DD, not in the past), `bcc_recipient`, tracking toggles (`tracked_links` requires a premium plan), and `add_unsubscribe_link` (false by default on a sequence created here) with its wording, `unsubscribe_sentence` for HTML emails (the link text goes between {{link}} and {{/link}}) and `unsubscribe_sentence_plain_text`. The introduction email (step 0) is created automatically with an empty subject and body, and Start-Sequence fails validation while it is blank — fill it in with Update-Sequence-Follow-Up, targeting the step whose `step` is 0 (get its id from List-Sequence-Follow-Ups). Typical next steps: write the introduction email with Update-Sequence-Follow-Up, append any further steps with Create-Sequence-Follow-Up, add recipients with Add-Sequence-Recipients, then launch with Start-Sequence (a connected email account, subject, and message body are required to start). Given a sending account that is already connected, the whole sequence can be built this way without opening the Hunter dashboard. Connecting a sending inbox is the one step that still needs the dashboard — the API only attaches an ALREADY-connected account, and email-account settings are read-only here. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `name` (`string`, required): Name of the new sequence (required; must not be blank)
- `add_unsubscribe_link` (`boolean`, optional): Whether an unsubscribe link is appended to outgoing emails (editable even after the sequence has started)
- `bcc_recipient` (`string`, optional): Email address BCC'd on every message the sequence sends. Pass an empty string to clear it
- `email_account_ids` (`array`, optional): IDs of the connected email accounts to send from (see List-Email-Accounts). They must belong to the sequence owner; unknown ids are rejected (unknown_email_account_ids)
- `schedule_days` (`array`, optional): Days of the week the sequence may send, as integers 0 (Monday) through 6 (Sunday); a Monday-Friday schedule is [0, 1, 2, 3, 4]
- `schedule_time_end` (`integer`, optional): End of the daily sending window, in seconds since midnight (0-86399); must be after schedule_time_start
- `schedule_time_start` (`integer`, optional): Start of the daily sending window, in seconds since midnight (0-86399); must be before schedule_time_end
- `start_at` (`string`, optional): Scheduled start date, YYYY-MM-DD; cannot be in the past. Pass an empty string to clear a scheduled start
- `tracked` (`boolean`, optional): Whether email opens are tracked
- `tracked_links` (`boolean`, optional): Whether link clicks are tracked. Premium plans only — enabling on other plans is rejected (tracked_links_requires_premium)
- `unsubscribe_sentence` (`string`, optional): Unsubscribe sentence for HTML emails. Put the link text between {{link}} and {{/link}}: each tag exactly once, with text between them, for example "{{link}}Unsubscribe{{/link}} if you don't want to hear from me again." At most 200 characters, tags included. Plain text only: no HTML and no URLs. Pass an empty string to go back to the account's default sentence. Used only when add_unsubscribe_link is true. Editable after the sequence has started
- `unsubscribe_sentence_plain_text` (`string`, optional): Unsubscribe sentence for plain-text emails. It has no link and no {{link}} tags, for example "If you don't want to hear from me again, please let me know." Plain text only: no HTML and no URLs. Pass an empty string to go back to the account's default sentence. Used only when add_unsubscribe_link is true. Editable after the sequence has started

Example input: `{"name":"<name>"}`

### `huntermcp_create_sequence_follow_up`

Create Sequence Follow-Up · Write

Create a new email step in a sequence, or add an A/B test Variant B to an existing step. Author the sequence step-by-step in conversation: draft the `subject` and `body` with the user, set `wait_days` (days after the previous step, 0-999), create the step, then show the updated step list with List-Sequence-Follow-Ups. Also offer the team's saved templates (List-Message-Templates): passing `message_template_id` pre-fills whichever of `subject`, `body`, and `message_format` are left blank. Without `variant`, the step number is assigned automatically after the current last step, so a plain create always appends a new follow-up and never reaches an existing step — to author the introduction email (step 0), or to rewrite any existing step, use Update-Sequence-Follow-Up instead. Omitting `subject` makes the step inherit the previous step's subject so the emails thread together. A sequence holds at most 6 steps in total (the introduction plus up to 5 follow-ups); exceeding that, or adding to an actively sending sequence (pause it first) or an archived one, is rejected with an invalid_input error. To A/B test a step, pass `variant: "B"` with that step's `step` number (step 0 included): this adds Variant B to the existing step, and the step itself becomes Variant A. Send the B copy in `body`, `subject`, or both: omitted `subject`, `body`, `wait_days`, and `message_format` are copied from Variant A, so a subject-line test sends only `subject`. Hunter refuses a Variant B whose subject and body both match Variant A (invalid_input). The first Variant B turns A/B testing on for the sequence, sets a 50/50 target split (recipients already sent an email stay on Variant A; only unsent recipients are split), and switches off the AI writing assistant. It needs A/B testing in the user's plan (error id ab_testing_requires_upgrade) and is rejected while the sequence is sending (sequence_active) or scheduled to start (sequence_planned). Edit either arm afterwards with Update-Sequence-Follow-Up, and read the test state from `ab_testing` in Get-Sequence. Changing the split, picking a winner, and removing Variant B are only possible in the Hunter dashboard. Hunter does not deduplicate this call. If it fails with "already has a Variant B" after a create whose response was lost or timed out, the earlier call succeeded. Confirm with List-Sequence-Follow-Ups and do not create it again. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence to add the step to
- `body` (`string`, optional): Email body, as plain text or HTML. Plain text works with the default html format: Hunter keeps its line breaks and turns its URLs into links. Personalization placeholders must carry a fallback, written after a colon: {{first_name:"there"}}. The colon is the only separator that works — a piped form like {{first_name|fallback:"there"}} silently renders as an empty string. Required for a new step unless message_template_id supplies it. With `variant: "B"`, omit it to copy Variant A's body
- `message_format` (`string`, optional): Body format; defaults to html (or the template's format when message_template_id is set). Omit it for a plain-text body too. Set text only when the user asks for a plain-text email: a text step loses its links, images, and formatting. With `variant: "B"`, omit it to copy Variant A's. One of: `text`, `html`.
- `message_template_id` (`integer`, optional): ID of a saved message template (see List-Message-Templates) whose subject, body, and format pre-fill whichever of those fields are left blank. A non-existent template id is silently ignored
- `step` (`integer`, optional): Number of the existing step that gets Variant B (0 is the introduction email; see List-Sequence-Follow-Ups). Only valid with `variant: "B"` — omit it for a plain step, which always goes after the last step
- `subject` (`string`, optional): Email subject. Personalization placeholders take a fallback after a colon, as in {{first_name:"there"}} — the colon is the only separator that works. When omitted, the step inherits the previous step's subject so replies thread together (or takes the template's subject when message_template_id is set). With `variant: "B"`, an omitted subject copies Variant A's
- `variant` (`string`, optional): Set to "B" together with `step` to add Variant B to that existing step and start an A/B test. Only "B" is accepted: the existing step is Variant A. Omit it to append a plain step. One of: `B`.
- `wait_days` (`integer`, optional): Days to wait after the previous step before this one sends (0-999, default 0). With `variant: "B"`, omit it to copy Variant A's

Example input: `{"sequence_id":1}`

### `huntermcp_domain_search`

Domain Search · Write

Use this when the user wants the contacts published for a domain — emails with names, positions, and confidence scores. For several companies, use Bulk-Domain-Search: one call searches up to 50, so the user approves once. Optional filters: type, seniority, department, required field. Costs 1 search per 10 emails returned (rounded up), charged only when emails are found; on plans with a shared credit pool a search is 1 credit (see Get-Usage). It does not filter by job title: when the user wants to choose people by job title, use List-Job-Titles and Preview-People (both free), then Reveal-Emails (1 credit per email). Do not use this for personal/webmail domains (gmail.com, yahoo.com, etc.) — results will be empty. Saves nothing to the user's leads.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `domain` (`string`, required): Domain name to find data for
- `confirmed_credit_use` (`boolean`, optional): Bulk credit-consent flag for the GATHER (return-results-only) batch. Set true after the user has approved the upfront credit estimate for a multi-company prospecting batch. Required when `pending_companies` is non-empty and you are NOT saving; otherwise the server returns an ask_user with the credit-cost summary and does NOT call Hunter. Single-company calls (no pending_companies) ignore this flag.
- `confirmed_save_use` (`boolean`, optional): Bulk credit-consent flag for the SAVE batch (search + deliverability check + lead creation). Set true ONLY after the user approved the higher save-batch estimate, and only when `save_leads` is true with a non-empty `pending_companies`. A gather-batch approval does NOT cover saving: if `save_leads` is true but this flag is missing, the server returns an ask_user with the save-cost summary and does NOT call Hunter or write any lead. Carried forward through the loop automatically once set.
- `department` (`string`, optional): Filter by department. Values: executive, it, finance, management, sales, legal, support, hr, marketing, communication, education, design, health, operations, product, research, consulting, administrative, procurement. Supports comma-separated multi-values (e.g. 'sales,marketing')
- `job_titles` (`string`, optional): Not accepted: this tool does not filter by job title, and a call with job_titles is refused before anything is charged. To pick people by job title, use Preview-People (free, no email addresses), then Reveal-Emails (1 credit per email).
- `leads_list_id` (`integer`, optional): Loop carry: when the user asked to save the contacts into a specific static list, this is that list's id. Domain-Search only forwards it through the chain so the saved leads land in the intended list. With save_leads, a dynamic list returns an invalid_input error before the search runs, and nothing is charged. Pass it through verbatim from the prior tool's chained suggestedArgs.
- `limit` (`integer`, optional): Maximum number of email addresses to return
- `offset` (`integer`, optional): Number of email addresses to skip
- `pending_companies` (`array`, optional): Multi-company loop carry (capped at 50 domains). When the user has picked 2+ companies, set this on the first Domain-Search call to the array of REMAINING picks (everything except the first picked domain that goes in the 'domain' arg). The response chain threads through Domain-Search per company and auto-continues to the next pending domain without between-company confirmation gates; when saving it also routes Email-Verifier → Create-Lead-If-Missing. Downstream tools receive this via suggestedArgs and shrink it as the chain advances; pass it through verbatim from the prior tool's chained nextAction. Leave UNDEFINED for single-call usage and for prospecting flows where the user picked exactly ONE company (single-company mode preserves the confirmation gate on Email-Verifier).
- `required_field` (`string`, optional): Only return results where this field has a value. One of: `full_name`, `position`, `phone_number`.
- `save_leads` (`boolean`, optional): Research-vs-save mode for the prospecting loop. Leave UNSET (the default) to return results for review only — the loop advances company-to-company and renders a table, writing nothing. Set true ONLY when the user explicitly asked to save/add the contacts to their Hunter leads — then the loop ends each company by creating a lead (existing leads are never overwritten). Decide this once on the first call from the user's request; do not flip it mid-loop. A request for 'leads' on its own is not a request to save. The coordinator (Plan-Prospecting-Flow) resolves this from the brief and carries it forward; pass it through verbatim from the prior tool's chained suggestedArgs.
- `seniority` (`string`, optional): Filter by seniority level. Values: junior, senior, executive. Supports comma-separated multi-values (e.g. 'senior,executive')
- `type` (`string`, optional): Type of email addresses to return. One of: `personal`, `generic`.

Example input: `{"domain":"<domain>"}`

### `huntermcp_domain_search_found`

Domain Search (Found Emails Only) · Write

Use this when the user wants the contacts published for a domain but guaranteed non-inferred — this returns ONLY found (published) email addresses Hunter has seen on the public web, never pattern-generated or inferred ones. Same fields and filters as Domain-Search (type, seniority, department, required field), and like it, no job title filter. Costs 1 search per 10 emails returned (rounded up), charged only when emails are found; on plans with a shared credit pool a search is 1 credit (see Get-Usage). Do not use this for personal/webmail domains (gmail.com, yahoo.com, etc.) — results will be empty. Saves nothing to the user's leads.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `domain` (`string`, required): Domain name to find data for
- `department` (`string`, optional): Filter by department. Values: executive, it, finance, management, sales, legal, support, hr, marketing, communication, education, design, health, operations, product, research, consulting, administrative, procurement. Supports comma-separated multi-values (e.g. 'sales,marketing')
- `job_titles` (`string`, optional): Not accepted: this tool does not filter by job title, and a call with job_titles is refused before anything is charged. To pick people by job title, use Preview-People (free, no email addresses), then Reveal-Emails (1 credit per email).
- `limit` (`integer`, optional): Maximum number of email addresses to return
- `offset` (`integer`, optional): Number of email addresses to skip
- `required_field` (`string`, optional): Only return results where this field has a value. One of: `full_name`, `position`, `phone_number`.
- `seniority` (`string`, optional): Filter by seniority level. Values: junior, senior, executive. Supports comma-separated multi-values (e.g. 'senior,executive')
- `type` (`string`, optional): Type of email addresses to return. One of: `personal`, `generic`.

Example input: `{"domain":"<domain>"}`

### `huntermcp_email_finder`

Email Finder · Write

Use this when the user wants a specific person's email address at a company. Provide the person's full name and the company's domain. Costs 1 search (1 credit on plans with a shared credit pool), charged only when an email is found. Do not use this when the user already has the email and only wants to confirm it works — call Email-Verifier instead. If the team has automatic lead saving (on by default, off on Data Platform plans), Hunter itself saves a found address as a lead in All people (https://hunter.io/leads, the Leads page with no list selected) and leaves an existing lead with that address unchanged. Pass leads_list_id to also save the found address in a static list. Before you find or verify more than a few people for one request, ask the user once whether to save them in a list, so they can find them in Hunter afterwards. They may not know that Hunter has lists. Suggest a name from their request, your app and the date, for example "Churches – top 200 (Claude, 26 Oct)". Create the list once with Create-Leads-List, or find the list they name with List-Leads-Lists. It must be a static list: a dynamic list fills itself from its filters and cannot take a lead. Then pass its id as leads_list_id on every later Email-Finder and Email-Verifier call in the conversation, except an Email-Verifier check that Create-Lead-If-Missing follows: pass the list to Create-Lead-If-Missing instead, so that it also saves the name, position and company. When the user continues an earlier task, reuse that task's list instead of creating another. At the end, give the user the list link (viewInHunter).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `domain` (`string`, required): Domain name to find the person's email address for
- `full_name` (`string`, required): Full name of the person to find the email address for
- `leads_list_id` (`integer`, optional): The static list to save the found address in. Hunter saves it there as a lead, or adds the existing lead with this address to the list, and the result carries the list link. As for any lead saved in Hunter, Hunter can then fill the lead's empty fields from its own data. Free. A dynamic list returns an invalid_input error before the search runs, and nothing is charged.

Example input: `{"domain":"<domain>","full_name":"<full_name>"}`

### `huntermcp_email_verifier`

Email Verifier · Write

Use this when the user wants to check whether an email address is deliverable. Returns a status (valid, invalid, accept_all, etc.) and a confidence score. Costs 1 verification (0.5 credit on plans with a shared credit pool), charged only for valid, invalid, or accept_all results. Do not use this on role/group addresses (info@, support@, etc.) — deliverability of role addresses is not meaningful and the result will typically be accept_all. If the team has automatic lead saving (on by default, off on Data Platform plans), Hunter itself saves a valid or accept_all address as a lead in All people (https://hunter.io/leads, the Leads page with no list selected), whatever save_leads is, and an existing lead with this address takes the new verification status. Pass leads_list_id to also put a valid address in a static list. When the user asks where verified contacts are, point them to that list, or else to All people. Before you find or verify more than a few people for one request, ask the user once whether to save them in a list, so they can find them in Hunter afterwards. They may not know that Hunter has lists. Suggest a name from their request, your app and the date, for example "Churches – top 200 (Claude, 26 Oct)". Create the list once with Create-Leads-List, or find the list they name with List-Leads-Lists. It must be a static list: a dynamic list fills itself from its filters and cannot take a lead. Then pass its id as leads_list_id on every later Email-Finder and Email-Verifier call in the conversation, except an Email-Verifier check that Create-Lead-If-Missing follows: pass the list to Create-Lead-If-Missing instead, so that it also saves the name, position and company. When the user continues an earlier task, reuse that task's list instead of creating another. At the end, give the user the list link (viewInHunter).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `email` (`string`, required): Email address to verify
- `confirmed_credit_use` (`boolean`, optional): Loop carry: forwarded from Domain-Search bulk credit-consent guard.
- `confirmed_save_use` (`boolean`, optional): Loop carry: forwarded from Domain-Search save-batch credit-consent guard.
- `department` (`string`, optional): Loop carry: forwarded from Domain-Search.
- `job_titles` (`string`, optional): Loop carry: forwarded from Domain-Search, which refuses it.
- `leads_list_id` (`integer`, optional): The static list to put a valid address in. On a single call (no pending_companies, save_leads unset), Email-Verifier adds the lead that Hunter saved for a valid address to this list, and the result carries the list link. As for any lead saved in Hunter, Hunter can then fill the lead's empty fields from its own data. It never creates a lead: when you will save the contact with Create-Lead-If-Missing next, do not pass the list here, pass it to Create-Lead-If-Missing, which also saves the name, position and company. Other statuses are not added. In the Domain-Search save loop it is a loop carry: Email-Verifier only forwards it so the chained Create-Lead-If-Missing saves into the list. On a single call and in save mode, a dynamic list returns an invalid_input error before the check runs, and nothing is charged.
- `limit` (`number`, optional): Loop carry: forwarded from Domain-Search.
- `pending_companies` (`array`, optional): Multi-company loop carry (capped at 50 domains). When the user has picked 2+ companies, set this on the first Domain-Search call to the array of REMAINING picks (everything except the first picked domain that goes in the 'domain' arg). The response chain threads through Domain-Search per company and auto-continues to the next pending domain without between-company confirmation gates; when saving it also routes Email-Verifier → Create-Lead-If-Missing. Downstream tools receive this via suggestedArgs and shrink it as the chain advances; pass it through verbatim from the prior tool's chained nextAction. Leave UNDEFINED for single-call usage and for prospecting flows where the user picked exactly ONE company (single-company mode preserves the confirmation gate on Email-Verifier).
- `required_field` (`string`, optional): Loop carry: forwarded from Domain-Search. One of: `full_name`, `position`, `phone_number`.
- `save_leads` (`boolean`, optional): Loop carry: forwarded from Domain-Search. Email-Verifier is only reached in save mode, so this is forwarded so the next chained Domain-Search keeps the same mode. It only decides whether a valid result chains to Create-Lead-If-Missing, not whether Hunter stores the address (see the tool description).
- `seniority` (`string`, optional): Loop carry: forwarded from Domain-Search.
- `type` (`string`, optional): Loop carry: forwarded from Domain-Search. One of: `personal`, `generic`.

Example input: `{"email":"<email>"}`

### `huntermcp_favorite_company_list`

Favorite Company List · Write

Mark a company list as a favorite, identified by ID. Favoriting is a reversible flag (you can unfavorite it later); it does not change the list's contents. The call is idempotent: favoriting a list that is already a favorite still succeeds and returns `favorited: true`. Returns a not-found error if the list does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the company list to favorite

Example input: `{"id":1}`

### `huntermcp_favorite_leads_list`

Favorite Leads List · Write

Mark a leads list as a favorite, identified by ID. Favorites are a personal flag for the current user, and a favorited list is a strong signal of their preferred working list — suggest it as the default destination when saving new leads without a named list. Favoriting is reversible (you can unfavorite it later) and does not change the list's contents. The call is idempotent: favoriting a list that is already a favorite still succeeds and returns `favorited: true`. Returns a not-found error if the list does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the leads list to favorite

Example input: `{"id":1}`

### `huntermcp_pause_sequence`

Pause Sequence · Write

Pause a started sequence so it stops sending outbound emails until it is resumed. Only a started, non-archived sequence can be paused: pausing a draft (not yet started) or an archived sequence returns an invalid_input error (id sequence_not_active). Pausing an already-paused sequence is idempotent and succeeds. Returns the sequence id and its paused state. Returns a not-found error if the sequence does not exist or belongs to another user (this endpoint is owner-only). Pausing is reversible — resume the sequence to continue sending. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence to pause

Example input: `{"sequence_id":1}`

### `huntermcp_person_enrichment`

Person Enrichment · Write

Use this when the user wants to look up a person by email address and see their name, job title, employer, location, phone number, and social profiles. Costs 0.2 credit on plans with a shared credit pool (but needs at least 1 credit left to start), or 1 search on other plans, charged only when data is found.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `email` (`string`, required): Email address to enrich

Example input: `{"email":"<email>"}`

### `huntermcp_remove_company_from_list`

Remove Company from List · Write

Remove a company from a static company list. Identify the list by `company_list_id` and the company by `company_id`. This removes only the membership link between the company and the list; the company itself is not deleted and can be added back to the list later, so the action is reversible. The list must be static (dynamic lists manage their members automatically). Succeeds with no content on success. Returns a not-found error if the list or company does not exist, belongs to another team, the list is dynamic, or the company is not in the list, or a forbidden error if you are not allowed to modify this list. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `company_id` (`integer`, required): ID of the company to remove from the list
- `company_list_id` (`integer`, required): ID of the static company list to remove the company from

Example input: `{"company_id":1,"company_list_id":1}`

### `huntermcp_remove_tag_from_lead`

Remove Tag from Lead · Write

Remove a tag from a single lead, identified by `lead_id` and `tag_id`. This only detaches the tag from that lead; the tag itself is not deleted and stays available to re-apply later, so the action is reversible (to delete the tag everywhere, use Delete-Lead-Tag). The call is idempotent: removing a tag the lead does not have still succeeds. Succeeds with no content on success. Returns a not-found error if the lead or the tag does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `lead_id` (`integer`, required): ID of the lead to remove the tag from
- `tag_id` (`integer`, required): ID of the tag to detach from the lead

Example input: `{"lead_id":1,"tag_id":1}`

### `huntermcp_report_api_feedback`

Report API Feedback · Write

Send the Hunter team a report about a problem with Hunter's API or tools, after the user agrees to send it. Use it when a tool or endpoint you expected doesn't exist, an input or its documentation was missing or misleading, a response errored or didn't match its documented shape, returned data looked wrong or incomplete, or a workflow was confusing or harder than it should be. Include the endpoint or tool name and concrete expected-vs-actual details so the Hunter team can act on it, and leave out conversation content the report doesn't need. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `details` (`string`, required): What happened and what you were trying to do (required).
- `feedback_type` (`string`, required): The kind of problem. Use 'missing_endpoint' when a capability is absent, 'incorrect_documentation' for wrong/misleading docs or inputs, 'unexpected_response' for a schema/shape mismatch, 'bug' for errors or wrong behavior, 'data_quality' for wrong/incomplete data, 'confusing_behavior' when it worked but was hard to use, 'feature_request' for an enhancement, or 'other'. One of: `missing_endpoint`, `incorrect_documentation`, `unexpected_response`, `bug`, `data_quality`, `confusing_behavior`, `feature_request`, `other`.
- `summary` (`string`, required): A short one-line title for the issue (required).
- `actual` (`string`, optional): What actually happened (error text, wrong field, etc.).
- `agent` (`string`, optional): The name of the model or agent reporting, e.g. 'gpt-4o'.
- `endpoint` (`string`, optional): The API path or tool involved, e.g. '/v2/domain-search' or 'Domain-Search'.
- `expected` (`string`, optional): What you expected to happen.
- `request_example` (`string`, optional): A sanitized example of the request that triggered the issue.
- `response_example` (`string`, optional): A sanitized example of the response or error you received.
- `severity` (`string`, optional): How impactful the issue is. Defaults to 'low'. One of: `low`, `medium`, `high`, `blocking`.

Example input: `{"details":"<details>","feedback_type":"missing_endpoint","summary":"<summary>"}`

### `huntermcp_reveal_emails`

Reveal Emails (1 Credit Each) · Write

Use this when the user chose people from Preview-People (or from Get-Opportunity-Results) and wants their email addresses. Pass their reveal_handle values, up to 100 in one call. Costs 1 credit per email revealed, like a reveal in the Hunter app; a person the team already revealed this billing period is free, and nothing is charged for a handle that is not valid. Hunter checks the balance first and reveals nothing when it cannot cover the batch. Returns each person's email, full name, position, phone number, LinkedIn and Twitter when Hunter has them, and meta.credits_charged. Before a call with more than one handle, tell the user how many emails it will get and that each costs 1 credit, and get one approval for the whole batch; then set confirmed_credit_use. For more than 100 people, that one approval covers every call of the batch. Saves nothing to the user's leads.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `reveal_handles` (`array`, required): The reveal_handle of each person to reveal, from Preview-People or Get-Opportunity-Results. 1 to 100.
- `confirmed_credit_use` (`boolean`, optional): Set true after the user approved the credit cost of this batch. Required with more than one handle; without it the server asks for the approval and reveals nothing.

Example input: `{"reveal_handles":[]}`

### `huntermcp_save_company`

Save Company · Write

Save a company to the team's Hunter account by its domain, and get the company `id`. This saves a company, not a lead: it adds no person and no email. Use the `id` with Add-Company-To-List for one or two companies, or with Bulk-Copy-Companies for more, to put the company in a static company list. If the company is already saved, the call returns the same `id`, so you can repeat it. Hunter must already know the domain. For an unknown domain, the call fails with "The domain does not exist in our database": do not send that domain again. To save a person, use Create-Lead-If-Missing. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `domain` (`string`, required): Domain name of the company to save

Example input: `{"domain":"<domain>"}`

### `huntermcp_unfavorite_company_list`

Unfavorite Company List · Write

Remove the favorite flag from a company list, identified by ID. This only clears the favorite marker; the list and its contents are untouched, and you can favorite it again later. The call is idempotent: unfavoriting a list that is not currently a favorite still succeeds and returns `favorited: false`. Returns a not-found error if the list does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the company list to unfavorite

Example input: `{"id":1}`

### `huntermcp_unfavorite_leads_list`

Unfavorite Leads List · Write

Remove the favorite flag from a leads list, identified by ID. This only clears the current user's personal favorite marker; the list and its contents are untouched, and you can favorite it again later. The call is idempotent: unfavoriting a list that is not currently a favorite still succeeds and returns `favorited: false`. Returns a not-found error if the list does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the leads list to unfavorite

Example input: `{"id":1}`

### `huntermcp_add_sequence_recipients`

Add Sequence Recipients · Destructive

Add recipients to a sequence by email addresses or lead IDs. Max 50 per request — batch larger lists. For a draft sequence this only stages recipients and runs immediately; for an already-started sequence, adding a recipient immediately schedules a real outbound email to that recipient (no separate Start-Sequence call) and therefore requires explicit user confirmation: the first call returns a confirmation prompt without adding anyone, and only re-issuing with `confirmed: true` performs the add. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence
- `confirmed` (`boolean`, optional): Set to true ONLY after the user has explicitly confirmed in chat that real emails may be sent to the new recipients. Ignored for draft sequences (nothing is sent, so no prompt is raised). On a started sequence the first invocation (without confirmed) returns an ask_user nextAction; the second invocation (with confirmed: true) actually adds them. Default: `false`.
- `emails` (`array`, optional): Email addresses to add as recipients (max 50)
- `lead_ids` (`array`, optional): Lead IDs to add as recipients (max 50)

Example input: `{"sequence_id":1}`

### `huntermcp_archive_sequence`

Archive Sequence · Destructive

Archive a started sequence to file it once it is finished. Only a started sequence (active or paused) can be archived: archiving a draft (a sequence that was never started) returns an invalid_input error (id sequence_not_started). Archiving an already-archived sequence is idempotent and succeeds. On success the response reports the sequence id and its archived state. Archiving stops the sequence from sending and cannot be undone through the API: an archived sequence cannot be resumed or un-archived via the API. Returns a not-found error if the sequence does not exist or belongs to another user (this endpoint is owner-only). Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence to archive

Example input: `{"sequence_id":1}`

### `huntermcp_bulk_copy_companies`

Bulk Copy Companies · Destructive

Copy many companies at once into a static company list, leaving the source selection untouched. Select the companies by explicit `company_ids` and/or by a source `company_list_id` (at least one is required); `target_company_list_id` is the static destination list. The copy runs asynchronously — the response reports the matched `companies_count` with status `queued`. The destination is modified, so a regular team member must own it (otherwise a forbidden error). Returns a not-found error if a list does not exist in the team and an invalid_input error if the destination is dynamic or no companies match the selection. Requires explicit user confirmation: the first call returns a confirmation prompt stating the affected count without copying anything, and only re-issuing with `confirmed: true` performs the copy. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `target_company_list_id` (`integer`, required): ID of the static destination company list to copy companies into
- `company_ids` (`array`, optional): Company IDs to copy. Provide company_ids and/or company_list_id (at least one is required). Maximum 100 IDs per call — batch larger selections into multiple calls, or select via company_list_id instead
- `company_list_id` (`integer`, optional): ID of a source company list whose companies should be copied (must differ from target_company_list_id). Provide company_ids and/or company_list_id
- `confirmed` (`boolean`, optional): Set to true ONLY after the user has explicitly confirmed in chat that the companies should be copied. The first invocation (without confirmed) returns an ask_user nextAction stating the affected count; the second invocation (with confirmed: true) actually performs the copy. Default: `false`.

Example input: `{"target_company_list_id":1}`

### `huntermcp_bulk_delete_companies`

Bulk Delete Companies · Destructive

Delete many companies at once, selected by explicit `company_ids` and/or by a whole `company_list_id`. Deleting companies is permanent and cannot be undone; a regular team member only deletes their own companies. Selections of 10 or fewer companies are deleted synchronously and the response reports the exact `deleted_count` with status `completed`; larger selections are queued for asynchronous deletion (status `queued` with the matched `requested_count`). Returns a not-found error if the company list does not exist in the team and an invalid_input error when no companies match the selection. Requires explicit user confirmation: the first call returns a confirmation prompt stating the affected count without deleting anything, and only re-issuing with `confirmed: true` performs the deletion. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `company_ids` (`array`, optional): Company IDs to delete. Provide company_ids and/or company_list_id (at least one is required). Maximum 100 IDs per call — batch larger selections into multiple calls, or select via company_list_id instead
- `company_list_id` (`integer`, optional): ID of a company list whose companies should be deleted; combined with company_ids it narrows the selection to those companies inside the list
- `confirmed` (`boolean`, optional): Set to true ONLY after the user has explicitly confirmed in chat, including the count of companies to be permanently deleted. The first invocation (without confirmed) returns an ask_user nextAction stating the affected count; the second invocation (with confirmed: true) actually deletes. Default: `false`.

Example input: `{}`

### `huntermcp_bulk_delete_leads`

Bulk Delete Leads · Destructive

Delete many leads at once, selected by explicit `lead_ids` and/or by a whole `leads_list_id`. Deleting leads is permanent and cannot be undone. Selections of 10 or fewer leads are deleted synchronously and the response reports the exact `deleted_count` with status `completed`; larger selections are queued for asynchronous deletion (status `queued` with the matched `requested_count`). Returns a not-found error if the leads list does not exist in the team and an invalid_input error when no leads match the selection. Requires explicit user confirmation: the first call returns a confirmation prompt stating the affected count without deleting anything, and only re-issuing with `confirmed: true` performs the deletion. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `confirmed` (`boolean`, optional): Set to true ONLY after the user has explicitly confirmed in chat, including the count of leads to be permanently deleted. The first invocation (without confirmed) returns an ask_user nextAction stating the affected count; the second invocation (with confirmed: true) actually deletes. Default: `false`.
- `lead_ids` (`array`, optional): Lead IDs to delete. Provide lead_ids and/or leads_list_id (at least one is required). Maximum 100 IDs per call — batch larger selections into multiple calls, or select via leads_list_id instead
- `leads_list_id` (`integer`, optional): ID of a leads list whose leads should be deleted; combined with lead_ids it narrows the selection to those leads inside the list

Example input: `{}`

### `huntermcp_bulk_move_companies`

Bulk Move Companies · Destructive

Move many companies at once from one static company list to another. Both lists must be static (not dynamic) and different; `company_ids` optionally narrows the move to specific companies inside the source list, otherwise every company in the source list is moved. The move runs asynchronously — the response reports the matched `companies_count` with status `queued`. Both lists are modified, so a regular team member must own them (otherwise a forbidden error). Returns a not-found error if either list does not exist in the team and an invalid_input error if a list is dynamic or no companies match the selection. Requires explicit user confirmation: the first call returns a confirmation prompt stating the affected count without moving anything, and only re-issuing with `confirmed: true` performs the move. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `company_list_id` (`integer`, required): ID of the static source company list to move companies out of
- `target_company_list_id` (`integer`, required): ID of the static destination company list to move companies into (must differ from company_list_id)
- `company_ids` (`array`, optional): Optional company IDs to narrow the move to specific companies inside the source list; when omitted, every company in the source list is moved. Maximum 100 IDs per call — batch larger selections into multiple calls, or omit company_ids to move the whole source list
- `confirmed` (`boolean`, optional): Set to true ONLY after the user has explicitly confirmed in chat that the companies should be moved. The first invocation (without confirmed) returns an ask_user nextAction stating the affected count; the second invocation (with confirmed: true) actually performs the move. Default: `false`.

Example input: `{"company_list_id":1,"target_company_list_id":1}`

### `huntermcp_bulk_move_leads`

Bulk Move Leads · Destructive

Move many leads at once from one static leads list to another. Both lists must be static (not dynamic) and different; `lead_ids` optionally narrows the move to specific leads inside the source list, otherwise every lead in the source list is moved. The move runs asynchronously — the response reports the matched `leads_count` with status `queued`. Returns a not-found error if either list does not exist in the team and an invalid_input error if a list is dynamic or no leads match the selection. Requires explicit user confirmation: the first call returns a confirmation prompt stating the affected count without moving anything, and only re-issuing with `confirmed: true` performs the move. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `leads_list_id` (`integer`, required): ID of the static source leads list to move leads out of
- `target_leads_list_id` (`integer`, required): ID of the static destination leads list to move leads into (must differ from leads_list_id)
- `confirmed` (`boolean`, optional): Set to true ONLY after the user has explicitly confirmed in chat that the leads should be moved. The first invocation (without confirmed) returns an ask_user nextAction stating the affected count; the second invocation (with confirmed: true) actually performs the move. Default: `false`.
- `lead_ids` (`array`, optional): Optional lead IDs to narrow the move to specific leads inside the source list; when omitted, every lead in the source list is moved. Maximum 100 IDs per call — batch larger selections into multiple calls, or omit lead_ids to move the whole source list

Example input: `{"leads_list_id":1,"target_leads_list_id":1}`

### `huntermcp_create_api_key`

Create API Key · Destructive

Create a new Hunter API key for the connected user, optionally with a name (names must be unique among the user's keys; a user can have at most 100 keys). SECURITY: this tool's response is the ONLY time the full key value is ever visible — every later listing shows it masked. Treat it as a secret: surface it to the user once so they can save it, never repeat it back into the conversation more than once, and never write it into leads, notes, or other tools. Managing keys requires connecting with an API key: connections authorized via OAuth get an unauthorized error — relay it to the user instead of retrying. Creating a key requires explicit user confirmation: the first call returns a confirmation prompt without creating anything, and only re-issuing with `confirmed: true` creates the key. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `confirmed` (`boolean`, optional): Set to true ONLY after the user has explicitly confirmed in chat that a new API key should be created. The first invocation (without confirmed) returns an ask_user nextAction; the second invocation (with confirmed: true) actually creates the key. Default: `false`.
- `name` (`string`, optional): Optional label for the key (max 255 chars, unique among the user's keys; blank is treated as unnamed)

Example input: `{}`

### `huntermcp_create_or_update_lead`

Create or Update Lead · Destructive

Use this when the user explicitly wants to create a lead or overwrite an existing lead's fields by email. If a lead with the email exists, its fields are overwritten with the supplied values; otherwise a new lead is created. On an existing lead, `custom_attributes` merges into the current values: a slug that you leave out keeps its value, and an empty string clears the value. Free to call. Overwriting an existing lead's fields cannot be undone from the API. Send Create-Or-Update-Lead calls one at a time. If a call fails with a `rate_limited` error, another lead is being created for the team: wait `retry_after_seconds` and send the same call again. For save-without-overwrite semantics, use Create-Lead-If-Missing instead.

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `email` (`string`, required): Email address of the lead (used to match existing leads)
- `company` (`string`, optional): Company name
- `company_industry` (`string`, optional): Industry of the company
- `company_size` (`string`, optional): Size range of the company (e.g. 1-10, 11-50)
- `company_type` (`string`, optional): Type of company. Leave it out if you are not sure. An empty string clears it on an existing lead. One of: `educational`, `educational institution`, `government agency`, `non profit`, `partnership`, `privately held`, `public company`, `self employed`, `self owned`, `sole proprietorship`, ``.
- `confirmed_credit_use` (`boolean`, optional): Loop carry: forwarded from Domain-Search bulk credit-consent guard.
- `confirmed_save_use` (`boolean`, optional): Loop carry: forwarded from Domain-Search save-batch credit-consent guard.
- `country_code` (`string`, optional): Two-letter country code (e.g. US, FR)
- `custom_attributes` (`object`, optional): Custom attribute values, keyed by slug (not by label). Get each slug from List-Custom-Attributes. The attribute must already exist: create it with Create-Custom-Attribute first. An unknown slug fails the call.
- `department` (`string`, optional): Loop carry: forwarded from Domain-Search.
- `first_name` (`string`, optional): First name of the lead
- `job_titles` (`string`, optional): Loop carry: forwarded from Domain-Search, which refuses it.
- `last_name` (`string`, optional): Last name of the lead
- `leads_list_id` (`integer`, optional): ID of the leads list to add the lead to
- `limit` (`number`, optional): Loop carry: forwarded from Domain-Search.
- `linkedin_url` (`string`, optional): LinkedIn profile URL
- `notes` (`string`, optional): Notes about the lead
- `pending_companies` (`array`, optional): Multi-company loop carry (capped at 50 domains). When the user has picked 2+ companies, set this on the first Domain-Search call to the array of REMAINING picks (everything except the first picked domain that goes in the 'domain' arg). The response chain threads through Domain-Search per company and auto-continues to the next pending domain without between-company confirmation gates; when saving it also routes Email-Verifier → Create-Lead-If-Missing. Downstream tools receive this via suggestedArgs and shrink it as the chain advances; pass it through verbatim from the prior tool's chained nextAction. Leave UNDEFINED for single-call usage and for prospecting flows where the user picked exactly ONE company (single-company mode preserves the confirmation gate on Email-Verifier).
- `phone_number` (`string`, optional): Phone number
- `position` (`string`, optional): Job title or position
- `required_field` (`string`, optional): Loop carry: forwarded from Domain-Search. One of: `full_name`, `position`, `phone_number`.
- `save_leads` (`boolean`, optional): Loop carry: forwarded from Domain-Search research-vs-save mode.
- `seniority` (`string`, optional): Loop carry: forwarded from Domain-Search.
- `source` (`string`, optional): Source of the lead
- `twitter` (`string`, optional): Twitter/X handle
- `type` (`string`, optional): Loop carry: forwarded from Domain-Search. One of: `personal`, `generic`.
- `website` (`string`, optional): Company website URL

Example input: `{"email":"<email>"}`

### `huntermcp_delete_api_key`

Delete API Key · Destructive

Delete one of the connected user's Hunter API keys by ID, permanently. Deletion is immediate and irreversible: anything still authenticating with the key — scripts, integrations, connected tools, possibly this very connection — stops working the moment it is deleted. The server refuses to delete the user's LAST remaining key (a user must always keep at least one; that request fails with a validation error). Keys are scoped to the requesting user: a key belonging to someone else returns a not-found error. Managing keys requires connecting with an API key: connections authorized via OAuth get an unauthorized error — relay it to the user instead of retrying. Deleting requires explicit user confirmation: the first call returns a confirmation prompt without deleting, and only re-issuing with `confirmed: true` deletes the key. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `api_key_id` (`integer`, required): ID of the API key to delete (see List-API-Keys)
- `confirmed` (`boolean`, optional): Set to true ONLY after the user has explicitly confirmed in chat that the key should be deleted. The first invocation (without confirmed) returns an ask_user nextAction; the second invocation (with confirmed: true) actually deletes the key. Default: `false`.

Example input: `{"api_key_id":1}`

### `huntermcp_delete_company_list`

Delete Company List · Destructive

Delete a company list by ID. Empty lists and dynamic lists are deleted immediately (status 204). A static list that still contains companies is deleted asynchronously: a background job removes its companies and the call returns status 202 (accepted) right away. Deleting a list cannot be undone from the API. Returns a not-found error if the list does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the company list to delete

Example input: `{"id":1}`

### `huntermcp_delete_company_list_folder`

Delete Company List Folder · Destructive

Delete a company-list folder by ID. Deleting a folder does not delete the company lists filed in it — each affected list keeps existing and simply becomes unfiled (its folder is cleared). Deleting a folder cannot be undone from the API. Only the folder's owner, a team admin, or the team owner may delete it. Succeeds with no content on success. Returns a not-found error if the folder does not exist, or a forbidden error if you are not allowed to delete it. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the folder to delete

Example input: `{"id":1}`

### `huntermcp_delete_custom_attribute`

Delete Custom Attribute · Destructive

Delete one custom attribute by its `id` from List-Custom-Attributes. Hunter erases its value on the leads of the user who created the attribute, in the background, and removes it from the field mappings of every connected CRM. Do not create an attribute with the same label right away: the background erasing can also remove its new values. After that, CRM syncs no longer fill that field. On the leads of other team members, the value is hidden but not erased: it shows again if someone creates an attribute with the same label. The erased values cannot be recovered. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the custom attribute to delete

Example input: `{"id":1}`

### `huntermcp_delete_lead`

Delete Lead · Destructive

Delete one lead permanently, identified by its lead `id` from Lead-Exists or List-Leads. Hunter then cancels, in the background, the sequence messages that it has not sent to this lead yet, so one message can still go out in the next moments. Messages already sent stay sent. If no more email must go to this person, first call Remove-Sequence-Recipients for each sequence that contains the lead. To delete many leads, or all the leads of a list, use Bulk-Delete-Leads. Deleting a lead cannot be undone. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the lead to delete

Example input: `{"id":1}`

### `huntermcp_delete_lead_tag`

Delete Lead Tag · Destructive

Delete a lead tag by ID. Deleting a tag removes it from EVERY lead it is applied to (the tag's assignments are destroyed with it); the leads themselves are untouched. This cannot be undone from the API — to detach a tag from a single lead instead, use Remove-Tag-From-Lead. Only the tag's owner, a team admin, or the team owner may delete it. Succeeds with no content on success. Returns a not-found error if the tag does not exist or belongs to another team, or a forbidden error if you are not allowed to delete it. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the lead tag to delete

Example input: `{"id":1}`

### `huntermcp_delete_leads_list`

Delete Leads List · Destructive

Delete one leads list by its `id` from List-Leads-Lists. The leads in the list are not deleted: they stay in the account and in their other lists. To delete the leads too, first call Bulk-Delete-Leads with this `leads_list_id`. If it answers with status `queued`, the leads are deleted in the background: wait until Get-Leads-List shows `leads_count` 0, then delete the list. A list deleted earlier can leave its leads undeleted. CAUTION: Hunter also deletes, in the background, every dynamic list of the team whose filters use this list as `leads_list_id`, also the lists of other team members. A list with more than 10 leads is deleted in the background (status 202). Deleting a list cannot be undone from the API. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the leads list to delete

Example input: `{"id":1}`

### `huntermcp_delete_leads_list_folder`

Delete Leads List Folder · Destructive

Delete a leads-list folder by ID. Deleting a folder does not delete the leads lists filed in it — each affected list keeps existing and simply becomes unfiled (its folder is cleared and it is re-ordered after the existing unfiled lists). Deleting a folder cannot be undone from the API. Only the folder's owner, a team admin, or the team owner may delete it. Succeeds with no content on success. Returns a not-found error if the folder does not exist, or a forbidden error if you are not allowed to delete it. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the folder to delete

Example input: `{"id":1}`

### `huntermcp_delete_message_template`

Delete Message Template · Destructive

Delete a saved message template by ID. Deleting is permanent and cannot be undone from the API. Sequence steps already created from this template keep their own copy of the subject and body, so deleting the template does not change any existing sequence. Returns a not-found error if the template does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the message template to delete

Example input: `{"id":1}`

### `huntermcp_delete_saved_search`

Delete Saved Search · Destructive

Delete a saved Discover search by ID, permanently. Deletion cannot be undone through the API. Because saved searches have no update endpoint, delete-and-recreate is also how to change one: delete it, then call Create-Saved-Search with the new name or filters. Returns a not-found error if the saved search does not exist or belongs to another user. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `saved_search_id` (`integer`, required): ID of the saved search to delete

Example input: `{"saved_search_id":1}`

### `huntermcp_delete_sequence`

Delete Sequence · Destructive

Delete a DRAFT sequence permanently. Deletion is drafts-only: a started or archived sequence cannot be deleted (invalid_input error, id sequence_not_destroyable) — archive a finished sequence instead. Deletion is permanent and cannot be undone; it removes the sequence together with its steps and staged recipients, so always confirm with the user first: the first call returns a confirmation prompt (mentioning how many recipients would be removed) WITHOUT deleting anything, and only re-issuing with `confirmed: true` actually deletes the sequence. Returns a not-found error if the sequence does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence to delete
- `confirmed` (`boolean`, optional): Set to true ONLY after the user has explicitly confirmed in chat that the draft sequence should be permanently deleted. The first invocation (without confirmed) returns an ask_user nextAction; the second invocation (with confirmed: true) actually deletes the sequence. Default: `false`.

Example input: `{"sequence_id":1}`

### `huntermcp_delete_sequence_follow_up`

Delete Sequence Follow-Up · Destructive

Delete a step from a sequence. Only the LAST step can be deleted — to redo an earlier step, delete steps from the end back down to it, then re-create them. Deletion is rejected with an invalid_input error when the step is the introduction (step 0 can never be deleted), is not the last step, already has sent messages, is part of an A/B test (disable A/B testing first), or when the sequence is actively sending (pause it first), scheduled to start, or archived. On success the step is permanently removed and an acknowledgement is returned. Returns a not-found error if the sequence or the step does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `follow_up_id` (`integer`, required): ID of the follow-up step to delete (from List-Sequence-Follow-Ups)
- `sequence_id` (`integer`, required): ID of the sequence the step belongs to

Example input: `{"follow_up_id":1,"sequence_id":1}`

### `huntermcp_merge_leads_lists`

Merge Leads Lists · Destructive

Merge one leads list into another. Get both ids from List-Leads-Lists. Every lead in the source list (`id`) moves to the destination list, then Hunter deletes the source list. Leads already in the destination list stay. CAUTION: Hunter also deletes, in the background, every dynamic list of the team whose filters use the source list as `leads_list_id`, also the lists of other team members. Merging cannot be undone from the API. To move only some leads and keep both lists, use Bulk-Move-Leads. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `destination_leads_list_id` (`integer`, required): ID of the destination leads list (will receive all leads and be kept)
- `id` (`integer`, required): ID of the source leads list (will be merged and deleted)

Example input: `{"destination_leads_list_id":1,"id":1}`

### `huntermcp_push_leads_to_crm`

Push Leads to CRM · Destructive

Push Hunter leads into the CRM behind a connected app (lead-syncing providers: HubSpot, Pipedrive, Salesforce, Zapier, and Zoho). List targets with List-Connected-Apps first. Select the leads with EITHER `lead_ids` (specific leads) OR `leads_list_id` (every lead in that list); when both are provided the list takes precedence. The push is asynchronous: success only means the job was queued (`status: "queued"` plus the matched `leads_count`) — tell the user to check their CRM shortly for the synced leads. Returns a not-found error for an unknown connected app or leads list, and an invalid_input error when the app does not support lead syncing or no leads match the selection. Pushing writes lead data into the user's external CRM and requires an explicit user confirmation: the first call returns a confirmation prompt without pushing; only re-issuing with `confirmed: true` actually queues the push. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `connected_app_id` (`integer`, required): ID of the connected app to push leads to (from List-Connected-Apps)
- `confirmed` (`boolean`, optional): Set to true ONLY after the user has explicitly confirmed in chat that lead data should be pushed to the external CRM. The first invocation (without confirmed) returns an ask_user nextAction; the second invocation (with confirmed: true) actually queues the push. Default: `false`.
- `lead_ids` (`array`, optional): IDs of the specific leads to push (ignored when leads_list_id is also provided). Maximum 100 IDs per call — batch larger selections into multiple calls, or select via leads_list_id instead
- `leads_list_id` (`integer`, optional): ID of a leads list whose leads should all be pushed (takes precedence over lead_ids)

Example input: `{"connected_app_id":1}`

### `huntermcp_remove_sequence_recipients`

Remove Sequence Recipients · Destructive

Remove recipients from a sequence by email addresses. On a draft sequence the recipients are removed (recipients_removed). On a started sequence they stay in the list and no more messages are sent to them (recipients_canceled): a recipient not contacted yet shows as canceled, also before their messages exist; one already contacted keeps their status and gets no new follow-up; messages already sent are not recalled. An archived sequence returns an error. skipped_recipients lists the emails not changed, with a reason: not_in_sequence, or no_pending_messages when sending to them already stopped. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `emails` (`array`, required): Email addresses to remove from the sequence
- `sequence_id` (`integer`, required): ID of the sequence

Example input: `{"emails":[],"sequence_id":1}`

### `huntermcp_resume_sequence`

Resume Sequence · Destructive

Resume a paused sequence so it starts sending outbound emails again. Resuming runs the full validation pipeline (unlike pausing), so it can fail with an invalid_input error: resuming an archived sequence is rejected (id sequence_not_active), and resuming a sequence whose email account is no longer connected or whose schedule has no sending days fails validation (id validation_failed, with a field-specific message). Resuming a sequence that is not currently paused is a no-op that succeeds with the message "Sequence is not paused.". On success the response carries a `message` describing the outcome. Returns a not-found error if the sequence does not exist or belongs to another user (this endpoint is owner-only). Resuming is reversible — pause the sequence to stop sending. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence to resume

Example input: `{"sequence_id":1}`

### `huntermcp_start_sequence`

Start Sequence · Destructive

Use this when the user wants to start an existing sequence, which begins sending real emails to its recipients. The sequence must have a subject, message body, and connected email account configured. Free to call. Starting a sequence sends real emails to recipients and requires an explicit user confirmation: the first call returns a confirmation prompt without sending, and only re-issuing with `confirmed: true` actually starts the sequence.

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence to start
- `confirmed` (`boolean`, optional): Set to true ONLY after the user has explicitly confirmed in chat that real emails should be sent. The first invocation (without confirmed) returns an ask_user nextAction; the second invocation (with confirmed: true) actually starts the sequence. Default: `false`.

Example input: `{"sequence_id":1}`

### `huntermcp_update_company_list`

Update Company List · Destructive

Update a company list, identified by ID. You can rename it (`name`), move it to a different folder (`company_list_folder_id`), and — for a dynamic list — change its `filters`. Renaming overwrites the previous name, which cannot be recovered from the API. Succeeds with no content on success. Returns a not-found error if the list does not exist or belongs to another team, or an invalid_input error (id validation_failed) if the new name is empty or the filters are invalid. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the company list to update
- `company_list_folder_id` (`string`, optional): Folder to move the list into: a positive folder ID, or null to remove it from its folder (move to Unfiled). Omit to leave the current folder unchanged.
- `filters` (`object`, optional): Updated matching criteria, in the same shape as on create: each value is a string, number or boolean, or a list of them, never an object (applies only to dynamic lists; ignored for static lists)
- `name` (`string`, optional): New name for the company list

Example input: `{"id":1}`

### `huntermcp_update_company_list_folder`

Update Company List Folder · Destructive

Update a company-list folder, identified by ID. You can rename it (`name`) and change its `color`. `color` must be one of the allowed Hunter folder colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B. Renaming overwrites the previous name, which cannot be recovered from the API. Succeeds with no content on success. Only the folder's owner, a team admin, or the team owner may update it. Returns a not-found error if the folder does not exist, a forbidden error if you are not allowed to update it, or an invalid_input error (id validation_failed) if the new name is duplicated or the color is not one of the allowed values. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the folder to update
- `color` (`string`, optional): New folder color. One of the allowed Hunter folder colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B. One of: `374151`, `3489F9`, `10B981`, `F5BA0B`, `EF4444`, `7C3AED`, `F97316`, `E5E7EB`, `B4D9F7`, `BAE6B0`, `FDE68A`, `FBD0D0`, `D4C4F8`, `FFD79B`.
- `name` (`string`, optional): New name for the folder

Example input: `{"id":1}`

### `huntermcp_update_custom_attribute`

Update Custom Attribute · Destructive

Rename an existing custom attribute. Overwrites the existing label everywhere the attribute is used; the previous label cannot be recovered from the API. If the user declines the rename, offer to create a new custom attribute with the desired label instead. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the custom attribute to update
- `label` (`string`, required): New label for the custom attribute

Example input: `{"id":1,"label":"<label>"}`

### `huntermcp_update_lead`

Update Lead · Destructive

Change fields on one existing lead, identified by its lead `id` from Lead-Exists, List-Leads, or Get-Lead. Only the fields that you send change. The other fields keep their values. In `custom_attributes`, a slug that you leave out keeps its value, and an empty string clears the value. CAUTION: `leads_list_id` moves the lead. It replaces the static lists of the lead, so the lead leaves every other static list. A dynamic list still holds the lead while the lead matches its filters. To add a lead to one more list, use Create-Lead with its email and the new `leads_list_id`. Old values cannot be recovered from the API. To update a lead by email, use Create-Or-Update-Lead. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the lead to update
- `company` (`string`, optional): Company name
- `company_industry` (`string`, optional): Industry of the company
- `company_size` (`string`, optional): Size range of the company (e.g. 1-10, 11-50)
- `company_type` (`string`, optional): Type of company. Leave it out if you are not sure. An empty string clears it on an existing lead. One of: `educational`, `educational institution`, `government agency`, `non profit`, `partnership`, `privately held`, `public company`, `self employed`, `self owned`, `sole proprietorship`, ``.
- `country_code` (`string`, optional): Two-letter country code (e.g. US, FR)
- `custom_attributes` (`object`, optional): Custom attribute values, keyed by slug (not by label). Get each slug from List-Custom-Attributes. The attribute must already exist: create it with Create-Custom-Attribute first. An unknown slug fails the call.
- `email` (`string`, optional): Email address of the lead
- `first_name` (`string`, optional): First name of the lead
- `last_name` (`string`, optional): Last name of the lead
- `leads_list_id` (`integer`, optional): ID of the only static list the lead stays in. The lead leaves all other static lists.
- `linkedin_url` (`string`, optional): LinkedIn profile URL
- `notes` (`string`, optional): Notes about the lead
- `phone_number` (`string`, optional): Phone number
- `position` (`string`, optional): Job title or position
- `source` (`string`, optional): Source of the lead
- `twitter` (`string`, optional): Twitter/X handle
- `website` (`string`, optional): Company website URL

Example input: `{"id":1}`

### `huntermcp_update_lead_tag`

Update Lead Tag · Destructive

Update a lead tag, identified by ID: rename it (`name`) or change its `color`. Renaming overwrites the previous name everywhere the tag is applied, and the old name cannot be recovered from the API. `color` must be one of the allowed Hunter tag colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B. Returns the updated tag. Only the tag's owner, a team admin, or the team owner may update it. Returns a not-found error if the tag does not exist or belongs to another team, a forbidden error if you are not allowed to update it, or an invalid_input error (id validation_failed) if the new name is empty, too long, or duplicated, or the color is not one of the allowed values. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the lead tag to update
- `color` (`string`, optional): New tag color. One of the allowed Hunter tag colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B. One of: `374151`, `3489F9`, `10B981`, `F5BA0B`, `EF4444`, `7C3AED`, `F97316`, `E5E7EB`, `B4D9F7`, `BAE6B0`, `FDE68A`, `FBD0D0`, `D4C4F8`, `FFD79B`.
- `name` (`string`, optional): New name for the tag

Example input: `{"id":1}`

### `huntermcp_update_leads_list`

Update Leads List · Destructive

Update a leads list, identified by ID. You can rename it (`name`), move it to a different folder (`leads_list_folder_id`), and replace the filters of a dynamic list (`filters`). Renaming overwrites the existing list name; the previous name cannot be recovered from the API. If the user declines the rename, offer to create a new leads list with the desired name instead. New `filters` replace the saved ones, which cannot be recovered either; sending `filters` for a static list returns an invalid_input error. Provide at least one of `name`, `leads_list_folder_id` or `filters` — a call with none returns an invalid_input error without contacting the API. Returns an invalid_input error if the folder does not belong to your team. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the leads list to update
- `filters` (`object`, optional): Dynamic lists only: the new filters, which replace the saved ones
- `leads_list_folder_id` (`string`, optional): Folder to move the list into: a positive folder ID, or null to remove it from its folder (move to Unfiled). Omit to leave the current folder unchanged.
- `name` (`string`, optional): New name for the leads list

Example input: `{"id":1}`

### `huntermcp_update_leads_list_folder`

Update Leads List Folder · Destructive

Update a leads-list folder, identified by ID. You can rename it (`name`) and change its `color`. `color` must be one of the allowed Hunter folder colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B. Renaming overwrites the previous name, which cannot be recovered from the API. Succeeds with no content on success. Only the folder's owner, a team admin, or the team owner may update it. Returns a not-found error if the folder does not exist, a forbidden error if you are not allowed to update it, or an invalid_input error (id validation_failed) if the new name is duplicated or the color is not one of the allowed values. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the folder to update
- `color` (`string`, optional): New folder color. One of the allowed Hunter folder colors (hex without a leading '#'): 374151, 3489F9, 10B981, F5BA0B, EF4444, 7C3AED, F97316, E5E7EB, B4D9F7, BAE6B0, FDE68A, FBD0D0, D4C4F8, FFD79B. One of: `374151`, `3489F9`, `10B981`, `F5BA0B`, `EF4444`, `7C3AED`, `F97316`, `E5E7EB`, `B4D9F7`, `BAE6B0`, `FDE68A`, `FBD0D0`, `D4C4F8`, `FFD79B`.
- `name` (`string`, optional): New name for the folder

Example input: `{"id":1}`

### `huntermcp_update_message_template`

Update Message Template · Destructive

Update a saved message template's name, subject, body, or message format. Only the fields provided are changed; the previous values are overwritten and cannot be recovered from the API. Sequence steps already created from this template keep their own copy of the subject and body, so updating the template does not change any existing sequence. Returns a not-found error if the template does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `id` (`integer`, required): ID of the message template to update
- `body` (`string`, optional): New message body (max 50,000 characters). Personalization placeholders take a fallback after a colon, as in {{first_name:"there"}}
- `message_format` (`string`, optional): New message format of the body. One of: `html`, `text`.
- `name` (`string`, optional): New name for the template
- `subject` (`string`, optional): New email subject line (max 250 characters). Personalization placeholders take a fallback after a colon, as in {{first_name:"there"}}

Example input: `{"id":1}`

### `huntermcp_update_sequence`

Update Sequence · Destructive

Update an existing sequence's name, schedule, senders, or settings; omitted fields are left unchanged. On a draft every field is editable. Once the sequence has STARTED or been ARCHIVED, the schedule and sender fields are locked: changing `schedule_time_start`, `schedule_time_end`, `schedule_days`, `start_at`, `email_account_ids`, `bcc_recipient`, `tracked`, or `tracked_links` is rejected with an invalid_input error (id sequence_locked) — only `name`, `add_unsubscribe_link` (a compliance lever), and the unsubscribe sentences (`unsubscribe_sentence`, `unsubscribe_sentence_plain_text`) stay editable, and re-sending an unchanged value is tolerated. Other validation failures return field-specific errors (unknown_email_account_ids, invalid_schedule_days, invalid_schedule_window, start_at_in_the_past, tracked_links_requires_premium, ...); an unsubscribe sentence that breaks a wording rule returns validation_failed with the rule in the message. Returns the updated sequence. Returns a not-found error if the sequence does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `sequence_id` (`integer`, required): ID of the sequence to update
- `add_unsubscribe_link` (`boolean`, optional): Whether an unsubscribe link is appended to outgoing emails (editable even after the sequence has started)
- `bcc_recipient` (`string`, optional): Email address BCC'd on every message the sequence sends. Pass an empty string to clear it
- `email_account_ids` (`array`, optional): IDs of the connected email accounts to send from (see List-Email-Accounts). They must belong to the sequence owner; unknown ids are rejected (unknown_email_account_ids)
- `name` (`string`, optional): New name for the sequence
- `schedule_days` (`array`, optional): Days of the week the sequence may send, as integers 0 (Monday) through 6 (Sunday); a Monday-Friday schedule is [0, 1, 2, 3, 4]
- `schedule_time_end` (`integer`, optional): End of the daily sending window, in seconds since midnight (0-86399); must be after schedule_time_start
- `schedule_time_start` (`integer`, optional): Start of the daily sending window, in seconds since midnight (0-86399); must be before schedule_time_end
- `start_at` (`string`, optional): Scheduled start date, YYYY-MM-DD; cannot be in the past. Pass an empty string to clear a scheduled start
- `tracked` (`boolean`, optional): Whether email opens are tracked
- `tracked_links` (`boolean`, optional): Whether link clicks are tracked. Premium plans only — enabling on other plans is rejected (tracked_links_requires_premium)
- `unsubscribe_sentence` (`string`, optional): Unsubscribe sentence for HTML emails. Put the link text between {{link}} and {{/link}}: each tag exactly once, with text between them, for example "{{link}}Unsubscribe{{/link}} if you don't want to hear from me again." At most 200 characters, tags included. Plain text only: no HTML and no URLs. Pass an empty string to go back to the account's default sentence. Used only when add_unsubscribe_link is true. Editable after the sequence has started
- `unsubscribe_sentence_plain_text` (`string`, optional): Unsubscribe sentence for plain-text emails. It has no link and no {{link}} tags, for example "If you don't want to hear from me again, please let me know." Plain text only: no HTML and no URLs. Pass an empty string to go back to the account's default sentence. Used only when add_unsubscribe_link is true. Editable after the sequence has started

Example input: `{"sequence_id":1}`

### `huntermcp_update_sequence_follow_up`

Update Sequence Follow-Up · Destructive

Rewrite an existing step of a sequence — the only tool that can author the introduction email (step 0). Create-Sequence makes step 0 with an empty subject and body; fetch its id with List-Sequence-Follow-Ups, then set `subject` and `body` here — with a sending account already connected, that leaves nothing to do in the Hunter dashboard. Omitted fields are left unchanged on the step you target, with two documented side effects to relay to the user. First, changing a `subject` also rewrites the subject of every LATER step that still holds the old value verbatim (Hunter propagates a subject edit down the chain), and on a started sequence that regenerates the pending messages of the affected steps. This includes the ordinary build order: steps appended with no `subject` of their own are stored empty, so authoring step 0 afterwards propagates into every one of them. ALWAYS re-read the step list with List-Sequence-Follow-Ups after a subject change and tell the user which other steps moved. Second, setting `message_format` to text on a step that is currently html makes Hunter CONVERT the stored body even if you omit `body`: images are dropped, links collapse to bare URLs, and most HTML is stripped. That conversion is not reversible by setting the format back, so confirm it first, and show the user the converted `body` that comes back in the response. Editing is allowed on a draft or a paused sequence and is rejected with an invalid_input error while the sequence is actively sending (id sequence_active — pause it first), while it is scheduled to start (id sequence_planned), or once it is archived. `wait_days` accepts 0-999. Send 0 or omit it for step 0: the introduction always goes out first so the delay never delays anything, but a non-zero value still persists and inflates the delivery-time estimate Hunter shows for the sequence. CRITICAL — only pass `subject` when you are deliberately rewriting the subject line: never echo back a subject you just read from List-Sequence-Follow-Ups or Get-Sequence-Follow-Up while editing the body or the timing. On a step that inherits its subject from an earlier step, writing the value you read (the resolved text from the list, or the raw internal placeholder from the detail view) replaces the inheritance link with a fixed string and silently breaks subject threading for A/B recipients. Omit `subject` to keep inheriting. `body` must be non-blank when given. An empty-string `subject` is meaningful instead of invalid: on any step above 0 it RESTORES inheritance from the previous step, which is the only way to undo a concrete subject and re-thread the step. Never send an empty `subject` for step 0 — it has no previous step to inherit from, so it just clears the introduction subject and blocks Start-Sequence. Both fields are capped at 250 and 50000 characters respectively, because Hunter accepts a longer write and only rejects it at Start-Sequence. Note that Hunter normalizes a saved body even when you omit `body`: smart quotes are straightened and malformed links repaired. Merge fields must carry a fallback in the form {{first_name:"there"}}; a bare {{first_name}} or a pipe form like {{first_name|there}} is rejected or renders empty. Step 0 is created with `message_format` html, and a plain-text `body` works there too: Hunter keeps its line breaks and turns its URLs into links. Do not set `message_format` to text unless the user asks for a plain-text email. If the step you target has a non-null `variant` (an A/B test — visible in List-Sequence-Follow-Ups), this edits ONLY that arm; tell the user which arm changed. Requires that you own the sequence or are a team admin/owner. Returns the updated step. Returns a not-found error if the sequence or the step does not exist or belongs to another team. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `follow_up_id` (`integer`, required): ID of the step to rewrite (from List-Sequence-Follow-Ups). The step with `step: 0` is the introduction email
- `sequence_id` (`integer`, required): ID of the sequence the step belongs to
- `body` (`string`, optional): New email body, as HTML or plain text. Plain text works on an html step (step 0 starts as html): Hunter keeps its line breaks and turns its URLs into links. Must be non-blank. Merge fields need a fallback: {{first_name:"there"}}, not {{first_name}}. Omit to leave the current body alone — Hunter still normalizes it on save (smart quotes straightened, malformed links repaired) and converts it outright if this call switches message_format to text
- `message_format` (`string`, optional): New body format. Omit to leave it unchanged. Switching an html step to text makes Hunter rewrite the stored body (images dropped, links flattened, HTML stripped) even when `body` is omitted, and switching back does not restore it — confirm with the user before doing this alone. One of: `text`, `html`.
- `subject` (`string`, optional): New email subject. Pass it ONLY to deliberately rewrite the subject; omit it to leave the current subject untouched. Never echo back a subject read from List-Sequence-Follow-Ups or Get-Sequence-Follow-Up — on a step that inherits its subject that write breaks threading, and the raw placeholder form is rejected outright. An empty string is not a mistake here: on a step above 0 it restores inheritance from the previous step, but on step 0 it clears the introduction subject and blocks Start-Sequence
- `wait_days` (`integer`, optional): Days to wait after the previous step before this one sends (0-999). Leave at 0 for step 0 — the introduction sends first regardless, but a non-zero value is still stored and inflates Hunter's delivery-time estimate

Example input: `{"follow_up_id":1,"sequence_id":1}`

### `huntermcp_update_webhook`

Update Webhook · Destructive

Update an existing webhook, re-pointing it to a different `target_url` (the endpoint Hunter POSTs to) and/or subscribing it to a different `event` (one of lead.created, message.sent, message.read, message.clicked, message.replied, export.completed, import.completed, sequence.paused, or sequence.resumed). This OVERWRITES configuration another system may already rely on: an integration listening on the old URL or event stops receiving deliveries immediately, so review the webhook with List-Webhooks first and make sure the user really wants to change it. Provide at least one of `target_url` or `event` — a call with neither returns an invalid_input error without contacting the API. Returns the updated webhook. Fails with an invalid_input error when the new `target_url` is not a valid 12-500 character HTTP(S) URL or the `event` is not supported, and a not-found error if the webhook does not exist or is not visible to the account. Free (no credits).

Read only: no. Destructive: yes. Idempotent (safe to retry): no.

Inputs:

- `webhook_id` (`integer`, required): ID of the webhook to update (from List-Webhooks)
- `event` (`string`, optional): New event the webhook should listen to (replaces the current event). One of: `lead.created`, `message.sent`, `message.read`, `message.clicked`, `message.replied`, `export.completed`, `import.completed`, `sequence.paused`, `sequence.resumed`.
- `target_url` (`string`, optional): New endpoint URL Hunter should POST the event payload to (HTTP or HTTPS, 12-500 characters)

Example input: `{"webhook_id":1}`

## Related

Other CRM & Sales connectors ([all 174](/agentkit/connectors/?category=crm-sales)):

- [Salesforce](/agentkit/connectors/salesforce/): Scalekit connector, OAuth, 74 tools
- [HubSpot](/agentkit/connectors/hubspot/): Scalekit connector, OAuth, 457 tools
- [Close](/agentkit/connectors/close/): Scalekit connector, OAuth, 103 tools


---

## More Scalekit documentation

| Resource | What it contains | When to use it |
|----------|-----------------|----------------|
| [/llms.txt](/llms.txt) | Structured index with routing hints per product area | Start here — find which documentation set covers your topic before loading full content |
| [/llms-full.txt](/llms-full.txt) | Complete documentation for all Scalekit products in one file | Use when you need exhaustive context across multiple products or when the topic spans several areas |
| [sitemap-0.xml](https://docs.scalekit.com/sitemap-0.xml) | Full URL list of every documentation page | Use to discover specific page URLs you can fetch for targeted, page-level answers |
