Name: Blujay
Author: blujayhq

Blujay MCP

Read and send texts over iMessage, RCS and SMS from your Blujay phone numbers.

Grade A Local · stdio API key 22 tools v1.0.3 MIT

No reviews yet by blujayhqnpm 887 Runs on your machine, nothing hosted

97/100

trust score

887

downloads a day

local

runs on your machine

tools declared

What it does

Read and send texts over iMessage, RCS and SMS from your Blujay phone numbers.

Quickstart

# 1 — add it to Claude Code
claude mcp add --scope user --env 'BLUJAY_API_KEY=${BLUJAY_API_KEY}' --transport stdio blujay-mcp '--' npx -y blujay-mcp

# 2 — ask your agent for something these tools do
#   get_account, list_conversations, get_conversation and 19 more

It needs BLUJAY_API_KEY: this reads it from your environment, so export it first.

Blujay costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.

Publishernot claimed

blujayhq1 server

1 category

All 1 server

Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.

Installation

Where are you running it?

Every route below installs the same thing and ends at the same approval screen. This one runs on your machine: your client starts Blujay as a process under your own user, with your files and your network, so the tool surface below is what it can do to your computer rather than to a server somewhere else. Its scan result is on the Security tab — read the surface before you approve it.

Claude Codedeclared by this listing

This is a public server: you run it yourself and this marketplace is not in the path. You need Node.js — the current LTS (24) is the safe choice; npx comes with it. Claude Code registers it in one command; --scope user makes it available in every project. Claude Code reads BLUJAY_API_KEY (Your Blujay API key, from API in the Blujay sidebar) from your environment each time it starts the server, so the value stays out of its config: export it in your shell profile (macOS, Linux), or run setx BLUJAY_API_KEY … and open a new terminal (Windows). The '--' is in quotes so that PowerShell passes it on.

claude mcp add --scope user --env 'BLUJAY_API_KEY=${BLUJAY_API_KEY}' --transport stdio blujay-mcp '--' npx -y blujay-mcp

Reconnect, or start a new session, and the tools appear in the model’s tool list.

Free on mcprush, and it runs from its publisher’s own package: this marketplace is not in the path, so there is no key of ours to hold and no card to add.

What it asks of you

It reaches a system of yours, so it needs your own credential rather than the publisher’s: BLUJAY_API_KEY — Your Blujay API key, from API in the Blujay sidebar.. You set it in your own client’s config, on your machine — this marketplace never holds it.

What it writes

One config entry your client uses to start the process locally. A local server runs with your file system and your network, which is why it is priced without metering.

What you approve

22 tools, with what each one reads, writes and reaches shown before you agree — the same list on every route above. Read the tool surface.

Tool surface

What the model actually sees. Descriptions are diffed on every release — see version history.

Only tools that write or reach the network

Read from its source code. A server that builds its tool list when it starts can list more tools than this.

Your Blujay workspace and its phone numbers: which number is which, whether each can send, and whether it can start conversations or only reply to people who texted first.

No parameter schema on file.

Your conversations, most recently active first. Filter to one person with with (their phone number or email) or to one of your numbers with on_number.

Parameter	Type	Required	Notes
with	string	no	Only conversations with this person: phone number or email.
on_number	string	no	Only conversations on this phone number of yours.
limit	integer	no	—
cursor	string	no	next_cursor from a previous call, for the next page.

The latest messages in one conversation as a transcript, oldest first, with delivery status on your messages. Pass person (phone number or email) or chat_id.

Parameter	Type	Required	Notes
person	string	no	Phone number or email of the person.
chat_id	string	no	—
limit	integer	no	How many recent messages.

Conversations where the other person sent the last message, so they're waiting on you.

Parameter	Type	Required	Notes
since_hours	integer	no	How far back to look.

One message's content and delivery: sent, delivered, read or failed, the service used, and why it failed.

Parameter	Type	Required	Notes
message_id	string	yes	—

Whether a phone number or email address uses iMessage, so you know if a text will arrive as iMessage or as RCS/SMS. An email address can only get iMessage.

Parameter	Type	Required	Notes
phone_number	string	yes	A phone number or an email address.

Search your contacts by name, phone number or email, or with none of those, list them (newest changes first).

Parameter	Type	Required	Notes
query	string	no	Name or part of one.
phone_number	string	no	—
email	string	no	—
unsaved	boolean	no	When listing: show the contacts nobody has saved yet instead of saved ones.
limit	integer	no	—
cursor	string	no	When listing: next_cursor from a previous call, for the next page.

Your webhook subscriptions. With subscription_id, also its most recent delivery attempts, to debug a webhook that isn't arriving.

Parameter	Type	Required	Notes
subscription_id	string	no	—

The workflows you can start for a contact with enroll_in_workflow: the active, published ones that start manually (or from the API).

No parameter schema on file.

Search docs.blujay.io for how something works: sending rules, webhooks, errors, limits, the SDKs.

Parameter	Type	Required	Notes
query	string	yes	—

Send a text from one of your Blujay numbers. This sends a real text to a real person; there is no test mode.

Parameter	Type	Required	Notes
to	array	no	Recipients: phone numbers, or email addresses (an email gets the text only over iMessage). Two or more start a group chat, which needs from.
chat_id	string	no	An existing conversation to send in.
text	string	no	—
attachment_url	string	no	Public https URL of a photo, video or contact card (vCard) to attach, up to 10 MiB.
file_path	string	no	Full path (or ~/...) of a photo, video, document (PDF, Word, Excel, PowerPoint), contact card (vCard) or audio file on this computer to send, up to 100 MiB. Never text files, code or keys. Only a path
voice_memo	boolean	no	Send attachment_url as a voice memo (it must be audio). Audio from file_path is always sent as a voice memo.
effect	string	no	iMessage effect for a text follow-up in an existing conversation (chat_id). Shows only over iMessage.
reply_to_message_id	string	no	Reply in the thread of this message. Only with chat_id.
from	string	no	Your phone number to send from, like +15555550100. Needed to start a group; otherwise usually leave it out.
send_again	boolean	no	Send even though the same text just went to the same place.

Add or remove a reaction on an iMessage: a tapback (love, like, dislike, laugh, emphasize, question) or any single emoji.

Parameter	Type	Required	Notes
message_id	string	yes	—
reaction	string	no	A tapback. Pass this or emoji.
emoji	string	no	Exactly one emoji, such as 😍 or 👍🏽, instead of a tapback. People on iOS 17 or earlier see it as a text.
remove	boolean	no	—

Change the text of an iMessage you sent, within 15 minutes and up to 5 times. The person sees the edit.

Parameter	Type	Required	Notes
message_id	string	yes	—
text	string	yes	—

Take back an iMessage you sent, within 2 minutes. It disappears for the person; this can't be undone.

Parameter	Type	Required	Notes
message_id	string	yes	—

Mark a conversation as read. On iMessage the person may see a read receipt.

Parameter	Type	Required	Notes
chat_id	string	yes	—

Show or hide the typing bubble in an iMessage conversation, for example while drafting a reply.

Parameter	Type	Required	Notes
chat_id	string	yes	—
typing	boolean	yes	—

Create a contact, or update the one with this phone number. Optionally add a tag. A tag can start a Blujay workflow that texts the contact.

Parameter	Type	Required	Notes
phone_number	string	yes	—
name	string	no	—
email	string	no	—
tag	string	no	—

Enroll one contact in a Blujay workflow, which starts its sequence of texts to them. This sends a real text to a real person; there is no test mode.

Parameter	Type	Required	Notes
workflow_id	string	yes	From list_workflows.
contact_id	string	yes	From find_contacts or save_contact.
request_id	string	no	Your ID for this enrollment, such as a signup or order ID. Sending it again returns the first result instead of enrolling twice.
metadata	object	no	Optional notes stored with the enrollment, such as its source. Up to 20 keys. Workflow steps can't read it and it doesn't change what's sent.

Stop all texts to a contact, for example after they ask you to stop. Only someone in the Blujay app can undo it.

Parameter	Type	Required	Notes
contact_id	string	yes	—
reason	string	no	—

Rename a group chat, or add or remove a person. Works only in iMessage groups, not SMS or RCS groups. Everyone in the group sees the change.

Parameter	Type	Required	Notes
chat_id	string	yes	—
rename_to	string	no	—
add_person	string	no	Phone number or email address to add.
remove_person	string	no	Phone number or email address to remove.

Send Blujay events (new messages, delivery updates, reactions, calls) to an HTTPS URL you run. Returns the signing secret once.

Parameter	Type	Required	Notes
url	string	yes	—
event_types	array	yes	For example ["message.received", "message.delivered", "message.failed"].

Call any Blujay API endpoint directly, for things the other tools don't cover. Look it up with search_docs first.

Parameter	Type	Required	Notes
method	string	yes	—
path	string	yes	—
body	object	no	—

Free

$0 forever

Every tool, no call limit
No card, and mcprush charges nothing for it
Source published under MIT
Runs on your machine — nothing of it reaches our gateway
Nothing to cap, because mcprush bills nothing

Nothing here is billable: mcprush charges nothing to install Blujay or to call it, at any volume.

ATrust score 97 / 100

Release 1.0.3 · scanned 12 Nov 2026

Two independent axes, because powerful and malicious are different questions. The grade is threat only. The capability level is blast radius, and it is never a penalty on the grade — it is priced as one subtract-only term in the score, where you can see it.

Threat score100

Client score97

Findings0 threat · 3 capability

Capabilitymoderate

Coveragesource

Verificationsource

Surface digest398f9253374b6b33f32678a5fe8c51b0098d9b07463fe0c1afe77b5fc30ee4c1

Score vectorthreat 100 → client 97

Term	Level	What it prices	Points
capability-exposure	moderate	capability blast radius (moderate) — client exposure if the model is manipulated	−3
verification-discount	source	publisher verification (provenance) — cryptographic build provenance ties the artifact to its source	−0
coverage-honesty	source	inspection depth (source) — how much of the target the scan could see	−0

score = clamp(round(threatScore − E_cap − E_ver − E_cov), 0, 100). No model, no opinion, no network call at scoring time: the same methodology version against the same bytes always produces this number, and every line above is a field of the report rather than a summary of it.

What the scan could actually read

A grade is only as meaningful as its coverage, so the scanner publishes its own depth before it publishes its result.

Tool surfaceread

Implementation sourceread

Package metadataread

Live transportnot available

Tools were statically extracted from the published source (22 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command "npx -y <package>".

11 of the 22 recovered tools are registered only when !options.readOnly holds (dist/server.js:472) — typically when a key or option is configured. They are part of the server's surface and were analysed; a server started without that configuration exposes only the other 10.

1 of the 22 recovered tools are registered only when options.enableInvokeApi holds (dist/server.js:900) — typically when a key or option is configured. They are part of the server's surface and were analysed; a server started without that configuration exposes only the other 10.

Capability — what it could do if the model were manipulated

Tags derived from each tool’s schema and the implementation, not from what the tool calls itself. moderate is the level these add up to.

10 tools are registered unconditionally · (1) 11 more only when !options.readOnly holds (dist/server.js:472) · (2) 1 more only when options.enableInvokeApi holds (dist/server.js:900) — typically when a key or option is configured. All 22 are in the table below and were analysed; the conditional ones are marked.

Tool	Capability tags	Why the tag was assigned
get_account	untrusted-input	untrusted-input: declares openWorldHint: true (interacts with entities outside the server)
list_conversations	untrusted-input	untrusted-input: declares openWorldHint: true (interacts with entities outside the server)
get_conversation	untrusted-input	untrusted-input: declares openWorldHint: true (interacts with entities outside the server)
find_unanswered	untrusted-input	untrusted-input: declares openWorldHint: true (interacts with entities outside the server)
get_message	untrusted-input	untrusted-input: declares openWorldHint: true (interacts with entities outside the server)
check_imessage	untrusted-input	untrusted-input: declares openWorldHint: true (interacts with entities outside the server)
find_contacts	untrusted-input	untrusted-input: declares openWorldHint: true (interacts with entities outside the server)
list_webhooks	untrusted-input	untrusted-input: declares openWorldHint: true (interacts with entities outside the server)
list_workflows	untrusted-input	untrusted-input: declares openWorldHint: true (interacts with entities outside the server)
search_docs	untrusted-input	untrusted-input: declares openWorldHint: true (interacts with entities outside the server)
send_message conditional (1)	external-sink	external-sink: keyword "send_message", declares readOnlyHint: false, openWorldHint: true
react_to_message conditional (1)	external-sink	external-sink: declares readOnlyHint: false, openWorldHint: true
edit_message conditional (1)	external-sink	external-sink: declares readOnlyHint: false, openWorldHint: true
unsend_message conditional (1)	external-sink	external-sink: declares readOnlyHint: false, destructiveHint: true, openWorldHint: true
mark_conversation_read conditional (1)	external-sink	external-sink: declares readOnlyHint: false, openWorldHint: true
set_typing conditional (1)	external-sink	external-sink: declares readOnlyHint: false, openWorldHint: true
save_contact conditional (1)	external-sink	external-sink: declares readOnlyHint: false, openWorldHint: true
enroll_in_workflow conditional (1)	external-sink	external-sink: declares readOnlyHint: false, openWorldHint: true
opt_out_contact conditional (1)	external-sink	external-sink: declares readOnlyHint: false, destructiveHint: true, openWorldHint: true
update_group conditional (1)	external-sink	external-sink: declares readOnlyHint: false, destructiveHint: true, openWorldHint: true
create_webhook conditional (1)	external-sink	external-sink: keyword "webhook", declares readOnlyHint: false, openWorldHint: true
invoke_api conditional (2)	external-sink	external-sink: declares readOnlyHint: false, destructiveHint: true, openWorldHint: true

Toxic-flow graph

The lethal trifecta, checked as a graph rather than as a checklist: untrusted input, a sensitive source and an external sink have to meet before there is a path worth worrying about.

flow-1

Untrusted input can drive an external action even though no sensitive source is exposed.

Supply chain and provenance

Integrityfirst-seen

Surface digest398f9253374b6b33

Scanned2026-11-12

This is the first scan of this surface here, so there is nothing yet to compare it against.

Every result on this tab comes from one deterministic pass over the published package — offline, rule by rule, and auditable line by line above. Same methodology version, same bytes, same score.