mattermost

Original：🇺🇸 English

Translated

Read and search Mattermost chat using the `mm` CLI. Use this skill whenever the user mentions Mattermost, chat messages, team chat, unread messages, DMs, channel history, mentions, or wants to catch up on what happened in chat. Also triggers when the user asks about specific people's messages, channel activity, searching for something someone said, or checking notifications. The CLI outputs agent-friendly JSON by default with thread IDs, bot detection, and channel refs - no parsing needed.

4installs

Sourcerhnvrm/mattermost-cli

Added on2026-05-01

NPX Install

npx skill4agent add rhnvrm/mattermost-cli mattermost

SKILL.md Content

View Translation Comparison →

Mattermost CLI (

mm

)

Read and search Mattermost from the command line. JSON output by default,

--human

for markdown.

Setup

Check if

mm

is already available:

bash

mm whoami

If that works, skip to "Start here" below.

If

mm

is not found, install it. Pick whichever works in your environment:

bash

pip install mattermost-cli    # adds `mm` to PATH
# or
pipx install mattermost-cli   # isolated global install
# or
uvx --from mattermost-cli mm  # run without installing (needs uvx)

Then authenticate (one-time):

bash

mm login --url https://your.mattermost.server
mm whoami   # verify

The login command will prompt for credentials interactively. If your server supports Personal Access Tokens (Profile > Security > Personal Access Tokens in the Mattermost UI), you can skip the prompt:

bash

mm login --url https://your.mattermost.server --token YOUR_TOKEN

For environment variables, multiple servers, and troubleshooting, see references/setup.md.

Start here:

mm overview

Always run this first. It returns mentions, unread channels, and active channels in a single call.

bash

mm overview              # last 6 hours (default)
mm overview --since 1d   # last 24 hours

The response has three sections:

mentions - posts that @-mention you, with root message context when it's a reply
unread - channels with unread messages, sorted by count
active_channels - channels with recent posts, sorted by recency

Each entry includes a

ref

field you can pass directly to other commands.

Reading messages

bash

mm messages <channel>                   # last 30 messages, chronological
mm messages <channel> --since 2h        # messages from last 2 hours
mm messages <channel> --threads         # thread index: root + reply count + last reply
mm messages @username                   # DMs with someone

<channel>

accepts a name (

off-topic

),

@username

for DMs, or a channel ID (for group DMs from overview output).

Threads

Every post includes a

thread_id

. Use it to read the full conversation:

bash

mm thread <thread_id>                   # root + last 9 replies
mm thread <thread_id> --limit 0         # entire thread
mm thread <thread_id> --since 1h        # just recent replies (root always included)

Searching and mentions

bash

mm search "deployment issue"
mm search "from:alice in:devops after:2025-01-01"
mm mentions                             # @-mentions in last 24h
mm mentions --since 3d

Mentions for replies include a

root

field with the original message, so you know what "this" or "it" refers to without a follow-up call.

Channel context

bash

mm channel <name>                       # purpose, header, member/pinned count
mm pinned <channel>                     # important/pinned posts
mm members <channel>                    # who's here + online status
mm channels --since 6h                  # all channels with recent activity
mm channels --type dm                   # just DMs

People

bash

mm user @someone                        # profile, role, status, timezone

Key JSON fields

Every post includes these fields so you can navigate without guesswork:

Field	What it's for
`thread_id`	Pass to `mm thread` to read full conversation
`ref`	On channel entries; pass to `mm messages`
`is_bot` / `bot_name`	Webhook and bot posts are flagged automatically
`root`	On reply-mentions; the original message being replied to
`is_reply` / `reply_count`	Thread structure
`reactions`	Emoji counts like `{"+1": 3, "white_check_mark": 1}`

Bot posts from webhooks automatically extract alert content from Slack-format attachments, so you see the actual alert text instead of empty messages.

mattermost

NPX Install

Tags

SKILL.md Content

Mattermost CLI (
`mm`
)

Setup

Start here:
`mm overview`

Reading messages

Threads

Searching and mentions

Channel context

People

Key JSON fields

Further reading

mattermost

NPX Install

Tags

SKILL.md Content

Mattermost CLI (mm)

Setup

Start here: mm overview

Reading messages

Threads

Searching and mentions

Channel context

People

Key JSON fields

Further reading

Mattermost CLI (
`mm`
)

Start here:
`mm overview`