# Bullpen (remote · bullpen.sale)

AI sales team that works the phone: gets a real person on the line. Agents sign up by paying (MPP).

- Trust score: 73/100 (medium)
- Change this week: +3
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-11-21

> **Recent critical change**: Authorization (2026-11-09). See the changelog below before you install this server.

## Components

- remote · `bullpen.sale`: 73/100 (this document), [markdown](https://verifymcp.io/servers/sale-bullpen-bullpen/bullpen.md), [page](https://verifymcp.io/servers/sale-bullpen-bullpen/bullpen)

## Channel facts

- Endpoint: `https://bullpen.sale/mcp`
- Transports: `streamable-http`
- Auth: `none`
- Version: `1.0.2`

## Trust breakdown

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

Scored 2026-11-21.

- **Endpoint Security**: 63/100
  - The endpoint's TLS certificate is valid, in date, and uses a strong key.
  - Authorisation check failed: no authorisation is required to call this server, and it exposes a tool marked destructive (delete_call).
  - HTTPS is enforced; there's no plaintext access path.
  - The HSTS (Strict-Transport-Security) header is present.
  - DNSSEC check failed: this domain isn't protected by DNSSEC.
- **Transport & Reachability**: 100/100
  - Verified streamable-http transport via a live MCP handshake.
- **Schema Quality & AI Usability**: 72/100
  - AI-judged instruction clarity (good).
  - Tool/resource definitions use about 3191 tokens (~75/item across 42 items; 42 tools + 0 resources), lean.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 53/100
  - Stability observed for 16 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 98/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 94% of tool parameters carry a description.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - All 4 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation.
  - An AI judge read all 43 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.

## Install

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

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

### Claude

```bash
claude mcp add --transport http sale-bullpen-bullpen 'https://bullpen.sale/mcp'
```

### Cursor

```json
{
  "mcpServers": {
    "sale-bullpen-bullpen": {
      "url": "https://bullpen.sale/mcp"
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "sale-bullpen-bullpen": {
      "type": "http",
      "url": "https://bullpen.sale/mcp"
    }
  }
}
```

### Codex

```toml
[mcp_servers.sale-bullpen-bullpen]
url = "https://bullpen.sale/mcp"
```

### opencode

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

### OpenClaw

```bash
openclaw mcp add sale-bullpen-bullpen --url 'https://bullpen.sale/mcp' --transport streamable-http
```

### Hermes

```yaml
mcp_servers:
  sale-bullpen-bullpen:
    url: "https://bullpen.sale/mcp"
```

### Netclaw

```json
{
  "McpServers": {
    "sale-bullpen-bullpen": {
      "Transport": "http",
      "Url": "https://bullpen.sale/mcp"
    }
  }
}
```

### Vellum

```bash
assistant mcp add sale-bullpen-bullpen -t streamable-http -u 'https://bullpen.sale/mcp'
```

### Other

```json
{
  "mcpServers": {
    "sale-bullpen-bullpen": {
      "type": "http",
      "url": "https://bullpen.sale/mcp"
    }
  }
}
```

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

## Changelog

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

### 2026-11-19 (score 73, +1)

No change was recorded against any check on this day. Stability & Change Management went from 43 to 47. That category is still filling its 30-day observation window: 13 days of observed history at the previous scan, 14 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-17 (score 72, +1)

No change was recorded against any check on this day. Stability & Change Management went from 37 to 40. That category is still filling its 30-day observation window: 11 days of observed history at the previous scan, 12 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-15 (score 71, +1)

No change was recorded against any check on this day. Stability & Change Management went from 30 to 33. That category is still filling its 30-day observation window: 9 days of observed history at the previous scan, 10 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-14 (score 70, 0)

- [security] Tool “create_campaign” rewrote its description, which is the text the model reads
- [security] Tool “find_companies” rewrote its description, which is the text the model reads
- [security] Tool “set_campaign_customs” rewrote its description, which is the text the model reads
- [security] Tool “set_mobile” rewrote its description, which is the text the model reads
- [security] Tool “work_campaign” rewrote its description, which is the text the model reads
- [functional regression] Schema quality: 2890 → 3191
- [functional] New tool “list_campaign_contacts”
- [cosmetic] “find_companies” added an optional parameter “campaign_id”
- [cosmetic] “work_campaign” added an optional parameter “contact_ids”
- [cosmetic] “work_campaign” added an optional parameter “limit”

### 2026-11-13 (score 70, +1)

No change was recorded against any check on this day. Stability & Change Management went from 23 to 27. That category is still filling its 30-day observation window: 7 days of observed history at the previous scan, 8 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-11 (score 69, +1)

No change was recorded against any check on this day. Stability & Change Management went from 17 to 20. That category is still filling its 30-day observation window: 5 days of observed history at the previous scan, 6 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-09 (score 68, −6)

- [critical regression] Authorization: partial → fail
- [security] New tool “delete_call”, which the server declares destructive

### 2026-11-08 (score 74, 0)

- [functional regression] Schema quality: 2289 → 2816
- [functional] New tool “draft_team_profile”
- [functional] New tool “get_team_profile”
- [functional] New tool “rename_crew_member”
- [functional] New tool “set_campaign_customs”
- [functional] New tool “set_team_profile”

## MCP tools (42)

### `bullpen_about` (~62 tokens)

What Bullpen is

What Bullpen does, what it costs, and how to get a workspace: free with $10 of credits (no card) when your person signs in, or by paying over MPP. Call this first if you have no key. (No API key needed.)

### `bullpen_signup` (~196 tokens)

Hire a bullpen (sign up by paying)

Start creating a Bullpen workspace for your principal by paying with the Machine Payments Protocol (MPP, Stripe). Returns the 402 payment challenge and the exact URL to retry with your MPP client (e.g. `npx @stripe/link-cli mpp pay <url>`). The paid retry returns an api_key; pass it to this MCP server as `Authorization: Bearer bp_live_…`. $25–$500, all of it becomes credits, plus $10 free. (No API key needed.)

Input parameters:

- `agent_name` (string): Your name, shown in the workspace log.
- `amount` (number): USD to load as credits, 25–500. Default 25.
- `company` (string, required): Company name for the workspace.
- `owner_email` (string, required): Email of the person who will own the workspace. They get an invite.
- `owner_name` (string): Owner's name, optional.

### `whoami` (~20 tokens)

Who am I

The workspace, member and role this API key acts as.

### `credits_status` (~26 tokens)

Credit balance

Prepaid credit balance and recent usage. Spending tools return a 402 error at zero.

### `top_up_credits` (~64 tokens)

Top up credits

Start an MPP credit top-up ($5–$500). Returns the 402 challenge and the URL to pay and retry with your MPP client, sending your key as X-Bullpen-Key.

Input parameters:

- `amount` (number, required): USD, 5–500.

### `ask_herman` (~80 tokens)

Ask Herman (runs the bullpen)

Talk to Herman, the manager agent who runs the whole bullpen in plain language: status, finding companies, running campaigns, sending the crew to call. Anything that reaches the outside world comes back as a pending action with an id — nothing goes out until confirm_herman_action.

Input parameters:

- `message` (string, required): What you want, in plain words.

### `confirm_herman_action` (~58 tokens)

Confirm or cancel Herman's pending action

Run or drop a pending action Herman proposed (a call, text or Salesforce update). Only confirm what your principal actually wants.

Input parameters:

- `decision` (string, required)
- `id` (string, required): Pending action id from ask_herman.

### `ask_brain` (~44 tokens)

Ask the company brain

Answer a question from the workspace's own knowledge (calls, notes, docs), with sources, limited by the key's role.

Input parameters:

- `question` (string, required): The question.

### `find_companies` (~106 tokens)

Find companies

Search for companies to sell to (e.g. 'HVAC companies in Salt Lake City'). Returns each company's id, name, main phone line, website, city and state. Pass campaign_id to add them to that campaign's contact list in the same call.

Input parameters:

- `campaign_id` (string): Optional: add the companies found to this campaign.
- `max` (number): Max results, default 20.
- `query` (string, required): Who you're looking for, with place.

### `research_company` (~34 tokens)

Research a company

Who to ask for at a company returned by find_companies.

Input parameters:

- `id` (string, required): Company id from find_companies.

### `get_a_human` (~120 tokens)

Get a human on the line

A crew member dials the number, works the phone menu, hold and AI receptionist until a real person answers, then connects the key's owner into the live call. Spends credits ($0.30 per crew minute; no extra charge when a person is reached). Company lines only unless the callee gave written consent.

Input parameters:

- `target` (string): Who to ask for.
- `to` (string, required): E.164 phone number, e.g. +18015551234.
- `why` (string): One line on why we're calling.

### `knock` (~82 tokens)

Knock on a company

Reach a human at a company's main line and leave a short hello pointing to your principal. Business hours, once a week per company. Spends credits.

Input parameters:

- `company` (string, required): Company name.
- `note` (string): What to leave.
- `target` (string): Who to ask for.
- `to` (string, required): Company main line, E.164.

### `send_text` (~43 tokens)

Send a text

Text from the workspace's line. Opt-outs are honored. Spends credits.

Input parameters:

- `body` (string, required): Message.
- `to` (string, required): E.164 number.

### `send_email` (~51 tokens)

Send an email

Email from the workspace's agent inbox.

Input parameters:

- `cc` (string): Optional cc.
- `subject` (string, required): Subject.
- `text` (string, required): Plain-text body.
- `to` (string, required): Recipient.

### `list_calls` (~109 tokens)

List recent calls

Recent calls, newest first: id, call SID, when, number, crew member, whether it was handed to your person, outcome, summary, and whether a transcript and recording exist. Reps see their own calls; managers and admins see all.

Input parameters:

- `limit` (number): How many, 1–50. Default 10.
- `since` (string): Optional ISO time; only calls after it.
- `to` (string): Optional phone number; only calls to or from it.

### `get_call_transcript` (~100 tokens)

Get a call's transcript and recording

One call in full: the transcript as text and as timed lines with speakers (crew, the person reached, and your person after a handoff), the summary, outcome, and a signed recording link (MP3, valid one hour). Pass an id or call SID from list_calls.

Input parameters:

- `id` (string, required): Call id or call SID (CA…) from list_calls.
- `recording` (boolean): Include a recording link. Default true.

### `delete_call` (~74 tokens)

Delete a call's record

Delete the transcript, summary and recording of one call, for good (a deletion request from the person called, or your team's retention policy). The call row, its outcome and its cost remain. Reps can delete their own calls; managers and admins any.

Input parameters:

- `id` (string, required): Call id or SID from list_calls.

### `floor_state` (~14 tokens)

Who's on the floor

Live and recent crew calls.

### `create_campaign` (~75 tokens)

Create a campaign

Create a calling campaign. It starts paused: add contacts (import_campaign_contacts, or find_companies with campaign_id), set its customs, then work_campaign switches it on.

Input parameters:

- `goal` (string): What a win looks like, e.g. 'book a 20-minute demo'.
- `name` (string, required): Campaign name.

### `import_campaign_contacts` (~49 tokens)

Import contacts

Add contacts to a campaign from CSV text (header row: company, phone, name, title, email).

Input parameters:

- `csv` (string, required): CSV text.
- `id` (string, required): Campaign id.

### `list_campaign_contacts` (~59 tokens)

Campaign contacts

Every contact in a campaign with its status (new, researched, excluded), phone, line type and whether it is already in a play.

Input parameters:

- `id` (string, required): Campaign id.
- `limit` (number): Max rows, default 200.

### `work_campaign` (~115 tokens)

Work a campaign

Send the crew to work a campaign now. With only the campaign id it takes the reachable contacts not yet in a play (company lines, not excluded), easiest first, up to limit (default 10, max 25), and switches a paused campaign on. Pass contact_ids to choose. Spends credits.

Input parameters:

- `contact_ids` (array): Optional: specific campaign contact ids.
- `id` (string, required): Campaign id.
- `limit` (integer): How many to start when no contact_ids are given (1-25).

### `onboarding_status` (~65 tokens)

What's left to set up

Checklist of what this workspace still needs before every feature works (Herman, compliance agreement, phone number, crew, your mobile, credits, Salesforce, Google, texting, team), in the order to do them, with the tool or link for each. Start here after signing up.

### `open_page` (~136 tokens)

Get a link to a Bullpen page

A link your person can open: floor (the live floor: who's on a call right now, and recent calls); human_line (Human line: your mobile and live bridge calls); herman (the app home, with Herman in the corner); campaigns (campaigns); companies (companies found and researched); integrations (phone numbers, texting registration, Salesforce and Google); billing (credits, billing and the Outbound Compliance Agreement); team (members and invites); api_keys (API keys and agent connections (revoke here)). "floor" is the live floor to watch calls. (No API key needed.)

Input parameters:

- `page` (string, required)

### `turn_on_herman` (~24 tokens)

Switch Herman on

Switch Herman on for this member so ask_herman works.

### `get_outbound_agreement` (~42 tokens)

Read the Outbound Compliance Agreement

The full Outbound Compliance Agreement and its version. Show it to your person before accept_outbound_agreement; calls and texts stay locked until it's accepted.

### `accept_outbound_agreement` (~101 tokens)

Accept the Outbound Compliance Agreement

Accept the agreement on your person's behalf ONLY after they have read it and told you they agree. Requires their full name as they typed it; recorded with this connection. Admin or manager only.

Input parameters:

- `full_name` (string, required): The accepting person's first and last name, as they typed it.
- `person_agreed` (boolean, required): true only if your person explicitly said they agree.
- `version` (string, required): Version from get_outbound_agreement.

### `get_team_profile` (~85 tokens)

Read the team profile

The team profile the crew works from: company name as spoken, what the team sells, trademarks and how to say them, tone, how calls open and close, words to use and never use, house rules, the knock line, caller-ID name, and per-campaign overrides. Read it first; if is_set is false, offer to build it with draft_team_profile.

### `draft_team_profile` (~102 tokens)

Draft the team profile from the website

Draft a team profile from the company's website and any notes from your person. Returns a draft and its sources; nothing is saved. Walk your person through it — name as spoken, product names, tone, words to use and avoid, house rules, the knock line — and save what they approve with set_team_profile.

Input parameters:

- `notes` (string): What your person said about how the crew should sound.
- `website` (string): Company website, if known.

### `set_team_profile` (~135 tokens)

Save the team profile

Save the team profile (admin only). Pass only fields your person approved: spoken_name, pronunciation, about, website, trademarks [{name, say_as}], tone, opening, closing, say [], never_say [], customs (house rules, prose), knock_line ({name} = crew member's first name, {team} = company), caller_name (≤15 chars). It shapes how the crew sounds; it cannot change the rules (the crew says it's an AI when asked, keeps to calling hours, honours do-not-call, invents nothing).

Input parameters:

- `profile` (object, required): Fields to set. Null clears a field.

### `set_campaign_customs` (~145 tokens)

Customs for one campaign

Override how one campaign sounds and who it is (admin only). Customs: tone, opening (said as the first line; {name} and {team} fill in), closing, say, never_say, customs, knock_line. Identity, for a campaign that calls as a different company: spoken_name, pronunciation, about, website, caller_name. Handoff: closer_name and handoff_number, the person and phone that ring when a live prospect wants a human on this campaign's calls (instead of the workspace's human line). Null returns a key to the team default.

Input parameters:

- `campaign_id` (string, required): Campaign id.
- `customs` (object, required)

### `rename_crew_member` (~124 tokens)

Rename a crew member

Rename a crew member for the team's culture or region, optionally with a new voice, persona or greeting. Admin only; paid workspaces (402 otherwise — buy_credits first). Keys from list_team's crew or onboarding_status.

Input parameters:

- `greeting` (string): Optional: their outbound opener.
- `key` (string, required): Crew member key, e.g. jasper.
- `name` (string, required): New first name.
- `persona` (string): Optional: who they are, a few lines.
- `voice_id` (string): ElevenLabs voice id, optional.

### `search_numbers` (~40 tokens)

Find phone numbers to buy

Available US numbers by area code, for the crew's lines.

Input parameters:

- `area_code` (string, required): 3-digit area code, e.g. 801.

### `buy_number` (~95 tokens)

Buy a phone number

Buy a number: $3/month from prepaid credits (first month charged now). Default is a crew line, and a crew member is put on it right away; for:"me" makes it your person's own caller ID instead. Admin or manager only. Confirm with your person before buying.

Input parameters:

- `e164` (string, required): Number from search_numbers, E.164.
- `for` (string): crew (default) or me.

### `hire_crew` (~39 tokens)

Hire the AI crew

Put an AI crew member (name, voice, role) on every free phone line so the crew can dial. Run after a number is bought.

### `set_mobile` (~109 tokens)

Set a mobile number

Set the mobile that rings when the crew reaches a person (get_a_human). This is your number for every call in the workspace; to send one campaign's live handoffs to someone else, set handoff_number with set_campaign_customs instead. Yours by default; managers can set a teammate's with user_id (from list_team).

Input parameters:

- `mobile` (string, required): E.164 or 10-digit US number.
- `user_id` (string): Teammate's user_id, optional.

### `buy_credits` (~57 tokens)

Card checkout for credits

A Stripe Checkout link your person opens to add credits by card ($20–$10,000). Agents with a wallet can use top_up_credits instead.

Input parameters:

- `usd` (number, required): Amount in USD, 20–10000.

### `connect_salesforce` (~32 tokens)

Connect Salesforce

A link your person opens to approve Bullpen in Salesforce, so every call writes there. Admin or manager only.

### `connect_google` (~27 tokens)

Connect Google

A link your person opens to approve Google Calendar and Gmail, for booking meetings and sending mail.

### `invite_teammate` (~60 tokens)

Invite a teammate

Invite someone to the workspace by email; they get a sign-in link that lands on the join page. Roles: rep, manager, admin (only admins invite admins).

Input parameters:

- `email` (string, required): Their email.
- `role` (string)

### `list_team` (~28 tokens)

List the team

Members (role, whether their mobile is set, Herman on) and pending invites with their links.

### `watch_companies` (~38 tokens)

Watch companies for signals

Add companies to Social Intel, which scans them for buying and risk signals.

Input parameters:

- `names` (string, required): Company names, comma or newline separated.

## Diagnostics

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

## Score history

- 2026-11-21: 73
- 2026-11-20: 73
- 2026-11-19: 73
- 2026-11-18: 72
- 2026-11-17: 72
- 2026-11-16: 71
- 2026-11-15: 71
- 2026-11-14: 70
- 2026-11-13: 70
- 2026-11-12: 69
- 2026-11-11: 69
- 2026-11-10: 68
- 2026-11-09: 68
- 2026-11-08: 74
- 2026-11-07: 74
- 2026-11-06: 73
- 2026-11-05: 73

## Common questions

### What is the Bullpen MCP server?

Bullpen is an MCP server listed in the public MCP registry as sale.bullpen/bullpen. AI sales team that works the phone: gets a real person on the line. Agents sign up by paying (MPP). This page covers its hosted endpoint (https://bullpen.sale/mcp).

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

Bullpen scores 73 out of 100 on VerifyMCP. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

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

Bullpen exposes 42 tools: bullpen_about, bullpen_signup, whoami, credits_status, top_up_credits, and 37 more. Their descriptions and schemas cost roughly 3,065 tokens of context every time the server is loaded.

### Does the Bullpen MCP server require authentication?

No. We connected to Bullpen without credentials and it answered, so anything it exposes is reachable by anyone who knows the address.

### Is the Bullpen MCP server still maintained?

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

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

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

## Links

- Remote endpoint: https://bullpen.sale/mcp
- Repository: https://github.com/GreatPyreneseDad/bullpen-mcp
- Website: https://bullpen.sale/
- Changelog RSS feed: https://verifymcp.io/servers/sale-bullpen-bullpen/bullpen.xml
- Changelog JSON feed: https://verifymcp.io/servers/sale-bullpen-bullpen/bullpen/changelog.json
- Trust report (JSON): https://verifymcp.io/servers/sale-bullpen-bullpen/bullpen.json
- HTML version of this page: https://verifymcp.io/servers/sale-bullpen-bullpen/bullpen
