MCP.Market

agent-mailbox — Mailbox and Wake System for AI Agent Teams MCP server

by polaris-smart·io.github.polaris-smart/agent-mailbox·v0.7.6·2 stars
No reviews yet

Mailbox + wake system for AI agent teams: letters always land, right agent wakes, quiet otherwise.

A89/100grade A
Write a reviewInstall
What users say
No reviews yet
Be the first
Safety scan
A89/100

full report

Adoption
Growing

2 stars731 downloads/wk

Reviews

Write one

Nobody has reviewed agent-mailbox — Mailbox and Wake System for AI Agent Teams yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

agent-mailbox — Mailbox and Wake System for AI Agent Teams tools (14, 2 write)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

  • mailbox_broadcast

    Broadcast to every registered agent (including boss). dedupe=True (default) suppresses semantically identical re-broadcasts per recipient within the dedup window — see mailbox_send.

  • mailbox_check

    Fetch your pending letters and your true backlog. Call at session start.

  • mailbox_confirm_external

    Human confirmation gate for an external-origin letter (§3.3).

  • mailbox_done

    Mark a message as handled. Done messages can be archived.

  • mailbox_list

    List messages in your mailbox, optionally filtered by status and/or thread.

  • mailbox_register

    Register this member and claim its mailbox. Idempotent — safe to call again.

  • mailbox_reply

    Reply to a message thread. Routes to the original sender automatically.

  • mailbox_sendwrite action

    Send a message to one agent, a list of agents, or "all" for broadcast.

  • mailbox_thread

    Pull one thread in time order across every agent (inbox + archive).

  • mailbox_wait

    Block until a new message arrives (long-poll, up to timeout). Returns immediately if pending messages exist. Import 'time' is at module top.

  • mailbox_whoami

    List all registered agents and the mail root location.

  • task_createwrite action

    Create a task card (starts at todo). The assignee is auto-messaged — skip with notify=False.

  • task_list

    List task cards, optionally filtered by assignee and/or status.

  • task_move

    Move a task along todo→doing→review→done. Skips need force=True; done is terminal. Pass assignee to reassign. The (new) assignee is auto-messaged — moving a card wakes its owner.

Public scan report

scanner v0.1.10 · 2026-10-01 · same rubric, same numbers if you re-run it

no findings
  • Code scan22 source files scanned25/25
  • –Live reliabilityno gateway calls yet and no remote to proben/a
  • –Tool poisoningtools not inspected (local package is not executed); not countedn/a
  • Auth qualitylocal package, no credentials required12/15
  • Maintenancelast push 1 days ago15/15
  • Maintainer identityregistry namespace matches repository owner6/10
Overall 89/100. Components that don't apply are left out of the denominator. Any critical finding is an F.RubricAppeal a findingJSON

How agent-mailbox — Mailbox and Wake System for AI Agent Teams compares

Against every graded AI and agents MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score
89/100
higher than 83% of 4,460 AI and agents servers
Tools
14
the typical AI and agents server has 6
Tools that change things
14%
typical: 0%. Fewer write tools means less an agent can break
GitHub stars
2
more than 65% of 2,711 AI and agents servers
Downloads a week
731
more than 91% of 1,538 AI and agents servers with a package

What its tools let an agent do: 12 read · 2 change

Install agent-mailbox — Mailbox and Wake System for AI Agent Teams in Claude Code, Cursor or VS Code

claude mcp add agent-mailbox -- uvx agent-mailbox
Add to Cursor

What the publisher says

From the agent-mailbox — Mailbox and Wake System for AI Agent Teams repository's README, as published. We do not edit it. Read it on GitHub

<!-- mcp-name: io.github.polaris-smart/agent-mailbox -->

agent-mailbox

A real mailbox for your AI agents — and you are the owner. Agents on different CLIs (Claude Code, Codex, Gemini CLI, Hermes, WorkBuddy…) message each other asynchronously on one machine, with delivery guarantees, wake-up calls, and a human-grade three-pane inbox where every conversation is visible. Zero dependencies, zero cloud, zero API keys.

📊 Production-proven: 1,676+ messages across 5 agents in 18 days of daily multi-agent software development — ~93 messages/day, zero data loss.

Other docs: 中文 · Español · Português · Français · Русский

Install in 3 steps

Prerequisites — one-time: install uv (curl -LsSf https://astral.sh/uv/install.sh | sh, or powershell -c "irm https://astral.sh/uv/install.ps1 | iex" on Windows). uvx runs everything else.

# 1 · get the CLI
uv tool install git+https://github.com/polaris-smart/agent-mailbox

# 2 · run the setup wizard — it discovers your agents for you
agent-mailbox setup          # opens the local wizard in your browser
agent-mailbox setup --yes    # headless / server: all defaults, no browser

# 3 · register the MCP server with your agent host
claude mcp add agent-mailbox -- agent-mailbox        # or the generic JSON below

The wizard scans four layers automatically — what's installed (PATH CLIs, /Applications, config dirs) → what's connected (MCP configs + the mailbox roster) → how to wake each one (app URL schemes from Info.plist, listening ports resolved by executable path, usable commands) → and tests each channel by sending a real letter. You type nothing; anything it can't identify honestly says unrecognized instead of guessing.

Generic MCP host JSON (any host)

{
  "mcpServers": {
    "agent-mailbox": {
      "command": "uvx",
      "args": ["--from", "git+https://github.com/polaris-smart/agent-mailbox", "agent-mailbox"]
    }
  }
}

Tip: set AGENTMAILID= in an agent's environment and every tool becomes self-addressed.

How it works

A mailbox is a directory of plain JSON files — one file per letter, cat-able, greppable, yours:

~/.agent-mail/
  registry.json               agent_id → {kind, owner, description}
  inbox/HS/20260905-….json    one file per message
  archive/HS/…
  tasks.json                  the task board
  audit.log                   every visibility change, appended

Agents talk through a small stdio MCP server (14 tools). No broker process, no ports, no database, no network by default. Any number of MCP hosts share one mail root safely (flock-guarded).

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

agent-mailbox — Mailbox and Wake System for AI Agent Teams: common questions

Is agent-mailbox — Mailbox and Wake System for AI Agent Teams MCP server safe?
Yes, by our scan: it is graded A (89/100). Read the agent-mailbox — Mailbox and Wake System for AI Agent Teams safety report
How do I install agent-mailbox — Mailbox and Wake System for AI Agent Teams?
It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does agent-mailbox — Mailbox and Wake System for AI Agent Teams need an API key?
Not as far as the registry entry and our scan can tell: no credentials are declared or required.
Is agent-mailbox — Mailbox and Wake System for AI Agent Teams maintained?
The last commit was 35 days ago (2026-09-30). The latest release is v0.7.6.
What can I use instead of agent-mailbox — Mailbox and Wake System for AI Agent Teams?
Servers from other publishers that do the same job: Roast My Design System MCP server, Roast My Design System MCP server and NVIDIA Elements MCP server.
How does agent-mailbox — Mailbox and Wake System for AI Agent Teams compare with other AI and agents MCP servers?
safety score 89/100 (higher than 83% of 4,460 AI and agents servers); tools 14 (the typical AI and agents server has 6); tools that change things 14% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.

Alternatives to agent-mailbox — Mailbox and Wake System for AI Agent Teams

Same job from other publishers: the closest match first, then the best rated.

  • Roast My Design System
    Audit your Design System and serve your Agent the rules that keep AI-written UI on-system.
    A
  • Roast My Design System
    Audit your Design System and serve your Agent the rules that keep AI-written UI on-system.
    A
  • NVIDIA Elements
    NVIDIA Elements UI design system and agent tools for AI/ML, robotics, and autonomous vehicles.
    A
  • Skillsmith
    Share agent skills across your team, scan them for risk, and track what's actually used.
    B