Do I need a Slack API token or admin access?

No. slack_search_messages reads Slack's locally-cached data on your Mac. There are no tokens to paste, no OAuth flow, and no admin app approval required.

Which Slack messages can it search?

It searches the channels and conversations cached locally by the Slack app on your Mac, using full-text substring matching. Content that isn't cached locally won't appear in results.

Is my Slack data sent to the cloud?

No. The search runs entirely on your Mac. Nothing is uploaded to LMCP or any third party. For web AIs, only the specific results you request pass through the local relay, and nothing is stored on a server.

Search Slack Messages with AI on Mac

The slack_search_messages tool lets your AI assistant search across your locally-cached Slack channels using full-text substring matching. Instead of clicking through Slack's search box and scrolling threads, you just ask. For example, type: "Search Slack for the message where Maria shared the staging deploy URL last week." The AI runs the search against your cached Slack data and hands back the matching messages with channel and sender context.

It works entirely on your Mac. There are no tokens to paste, no Slack admin app to authorize, and nothing leaves your machine. LMCP is a free, native macOS MCP server with 150+ tools that connect your AI to the real apps you already use.

Which AI agents work?

Desktop AI clients auto-configure through local stdio, so the tool just appears after install and restart:

Claude Desktop
Cursor
VS Code (GitHub Copilot)
Windsurf
Zed

Web-based AIs connect through the LMCP Cloud Relay connector, which bridges them to your local Mac:

ChatGPT
Claude.ai (web)
Grok
Perplexity

Automation

Searching is rarely the end goal — it's the first step in a chain. Because the AI has 150+ tools at once, it can act on what it finds. Ask it to "find the Slack message about the client demo and add the date to my calendar," and it chains slack_search_messages with your Calendar tools. Or "search Slack for the bug report from QA and draft a reply in Mail" — it pulls the thread, then composes an email. The search result becomes input for Reminders, Notes, Messages, or any other domain.

Context

The AI works against your real Slack history, not a generic web answer. It searches the channels and DMs cached on your Mac, so when you ask about "that decision in #product last sprint," it surfaces the actual conversation — names, timestamps, and the exact wording. That grounding turns vague questions into precise, sourced answers instead of guesses.

Productivity

Slack search is fast at finding a keyword but slow at synthesizing. With the AI driving slack_search_messages, you skip the scroll-and-skim ritual: ask one question, get the relevant messages summarized, and move on. A five-minute hunt across three channels for a link, a decision, or who said what becomes a single sentence. Over a week of standups, threads, and pings, that adds up to real time back.

Privacy & GDPR

Everything runs locally on your Mac. The tool reads Slack's own locally-cached data — no Slack API token, no server-side copy, nothing uploaded. Your messages are never sent to LMCP or any third party. This is privacy by architecture, not by policy. Learn more about how this design keeps you compliant: GDPR-compliant by architecture.

Ready to try it? Download LMCP, drag it to Applications, and ask your AI to search Slack.