feat: extract email metadata from file-type messages

[derodero24]

Problem

When emails are forwarded to a Slack channel (via Slack's email integration), the message content is stored in the files[] array with filetype: "email" rather than in text or blocks. This results in completely empty text output from conversations_history and conversations_search tools.

Example Slack API response for an email message:

{
  "text": "",
  "blocks": [],
  "attachments": [],
  "files": [{
    "filetype": "email",
    "mode": "email",
    "subject": "Meeting Tomorrow",
    "plain_text": "Hi team, please join the meeting at 3pm.",
    "from": [{"name": "John Doe", "address": "john@example.com"}],
    "to": [{"name": "Email", "original": "Email <channel@workspace.slack.com>"}]
  }]
}

The MCP tool currently outputs an empty text column for these messages, even though the email subject and sender information is available in the files[] metadata.

Proposed Solution

Extract email metadata (subject, from) from files[] when:

• msg.Text is empty
• BlocksToText(msg.Blocks) returns empty (no Block Kit content)
• The message contains files with filetype == "email"

This would produce output like: [Email] From: John Doe (john@example.com) Subject: Meeting Tomorrow

Limitations

The plain_text field (email body) is not mapped in the slack-go/slack library's File struct, so only metadata (subject, sender) can be extracted via the standard API. Full email body content can still be retrieved via the existing attachment_get_data tool using the file ID.

Related

• Feature Request: Extract text content from message blocks in conversations_history #186 — Block Kit text extraction (addresses a different subset of empty-text messages)

[NasAndNora]

Bumping this — I'm hitting the exact same issue. Forwarding Gmail to a Slack channel via Slack's native email integration, and the MCP returns empty text for all email messages. Would love to see email metadata extraction (subject, from, body) landed. Happy to test if a PR is opened.

[derodero24]

#192 addresses the metadata portion of this issue (From, CC, Subject extraction).

Email body (plain_text) could not be extracted because slack-go/slack's File struct didn't map that field. I submitted slack-go/slack#1522 to add PlainText and PreviewPlainText fields, and it was merged today.

Once slack-go cuts a release containing that change (currently unreleased, after v0.18.0-rc2), we can update the dependency and add body extraction to FilesToText().

Keeping this issue open to track the remaining work.