mcp-notify
A minimal MCP server for sending Slack messages, supporting both bot and user tokens with concurrent safety and an AI bot signature in user-token mode.
README
mcp-notify
Minimal MCP server for sending Slack messages. Supports both bot tokens (xoxb) and user tokens (xoxc + cookie auth). Fire-and-forget — single send_message tool, nothing else.
Looking for a full-featured Slack MCP? Check out korotovsky/slack-mcp-server — it supports reading, searching, reactions, threads, DMs, and much more. We recommend it for most use cases.
This project exists because we needed two things it doesn't offer:
- Bot signature — in user-token mode, every message gets a
:robot_face:context block so it's clear the message was sent by an AI assistant, not a human. In bot-token mode, the app name/icon already serves this purpose.- Concurrent safety — SQLite with WAL mode and check-lock-recheck sync pattern, safe for multiple MCP instances running in parallel
Installation
Claude Code
Add to your MCP config (~/.claude.json globally or .mcp.json per project).
Bot token (recommended for Slack apps):
{
"mcpServers": {
"mcp-notify": {
"type": "stdio",
"command": "npx",
"args": ["-y", "@pilat/mcp-notify"],
"env": {
"SLACK_MCP_BOT_TOKEN": "xoxb-..."
}
}
}
}
User token (xoxc + cookie):
{
"mcpServers": {
"mcp-notify": {
"type": "stdio",
"command": "npx",
"args": ["-y", "@pilat/mcp-notify"],
"env": {
"SLACK_MCP_XOXC_TOKEN": "xoxc-...",
"SLACK_MCP_XOXD_TOKEN": "xoxd-..."
}
}
}
}
Other MCP clients
Use npx @pilat/mcp-notify as the command with stdio transport. Pass credentials as environment variables:
# Bot token
SLACK_MCP_BOT_TOKEN=xoxb-... npx @pilat/mcp-notify
# User token
SLACK_MCP_XOXC_TOKEN=xoxc-... SLACK_MCP_XOXD_TOKEN=xoxd-... npx @pilat/mcp-notify
Environment Variables
| Variable | Required | Description |
|---|---|---|
SLACK_MCP_BOT_TOKEN |
One of bot/user | Bot token (xoxb-...) from your Slack app |
SLACK_MCP_XOXC_TOKEN |
One of bot/user | User's xoxc-... token |
SLACK_MCP_XOXD_TOKEN |
With xoxc | User's xoxd-... session token (value of the d cookie) |
SLACK_MCP_DATA_DIR |
No | Custom path for SQLite cache (default: ~/.local/share/mcp-notify) |
Set either SLACK_MCP_BOT_TOKEN or SLACK_MCP_XOXC_TOKEN + SLACK_MCP_XOXD_TOKEN. Setting both bot and user tokens is an error.
Bot token setup
- Create a Slack app at api.slack.com/apps
- Add the following OAuth scopes:
channels:read,groups:read,users:read,usergroups:read,chat:write - Install the app to your workspace
- Copy the Bot User OAuth Token (
xoxb-...)
User token setup (xoxc + cookie)
- Open Slack in browser (not desktop app)
- Open DevTools → Network tab
- Make any action in Slack (switch channel, send message)
- Find any request to
api.slack.com→ Headers tab - SLACK_MCP_XOXC_TOKEN: from request payload, find
token=xoxc-... - SLACK_MCP_XOXD_TOKEN: from Cookie header, find
d=xoxd-...— copy only thexoxd-...part (withoutd=)
Architecture
Single tool: send_message. In user-token mode, messages include a :robot_face: context block. In bot-token mode, messages are sent without the signature (the app name/icon already identifies the sender). Both modes use Block Kit with plain text fallback.
- SQLite cache (
~/.local/share/mcp-notify/data.db, WAL mode) — channels, users, user groups with lazy sync on first cache miss, 24h TTL - Mention resolution —
@username→<@U123>,@grouphandle→<!subteam^ID>. Groups take priority. Resolved in parallel. - Concurrent sync safety — CAS-based check-lock-recheck pattern via
sync_metatable
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.