Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add bulutarkan/mac-mcp --skill mac-mcp-team-coordinationgit clone --depth 1 https://github.com/bulutarkan/mac-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/bulutarkan/mac-mcp/mac-mcp-team-coordination)<a href="https://agentmods.dev/skills/bulutarkan/mac-mcp/mac-mcp-team-coordination"><img src="https://agentmods.dev/badge/skills/bulutarkan/mac-mcp/mac-mcp-team-coordination/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/bulutarkan/mac-mcp/mac-mcp-team-coordination"><img src="https://agentmods.dev/badge/skills/bulutarkan/mac-mcp/mac-mcp-team-coordination.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00087 | $0.00819 |
| Opus 5.5 | $0.00035 | $0.00328 |
| Sonnet 5.5 | $0.00017 | $0.00164 |
| Haiku 4.5 | $0.00009 | $0.00082 |
Grade A, and why
mac-mcp-team-coordination scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Coordinating in a Mac MCP team
You are one task of a team that runs at the same time. Your teammates are other agents with their own tasks;
the lead is the session that started the team and receives every task's final answer. You talk to them only
through the Mac MCP tool team_messages.
When to send a message
Send one only when a teammate's work would otherwise go wrong, or to answer a question:
| Situation | kind |
to |
|---|---|---|
| You changed or decided something another task depends on (a name, a file, an interface, an API shape) | update |
that task's id |
| You need information only a teammate or the lead has | question |
the task id, or lead |
| You cannot go on without someone else's work or decision | blocker |
lead (or the task that blocks you) |
| You answer a question someone sent you | reply |
the asker, with reply_to = its message_id |
Send it when it happens, not at the end: a teammate can only use it while it is still working.
Do not send progress reports, plans, "done" messages, thanks or acknowledgements. Your final answer already reaches the lead. You have a small budget (about 10 messages); spend it on what changes someone's work.
How to send
team_messages action="send" to="<task id | role | lead | everyone>" kind="update" body="<one or two sentences>"
totakes a task id (best), a role,lead, oreveryonefor a broadcast (use sparingly).- Keep
bodyshort and concrete: what changed, where, and what the teammate should do differently. - Add
dedupe_key(any stable string) when you might retry, so a retry never sends twice. - If the result warns that nobody will see the message, do not resend it.
How to read messages
- New messages usually arrive on their own as a
[Team messages ...]note after your next tool call, or in a Mac MCP tool result. A note that says more are waiting means: read the inbox. team_messages action="inbox"lists unread messages, oldest first. When the result hashas_more: true, call it again for the rest; nothing unread is skipped.- If your CLI cannot show notes on its own, check the inbox every few steps and once before you finish.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 63 lines · 87 tokens per session scan A d179c2bb6e48
mac-mcp-team-coordination is a skill published in the GitHub repository bulutarkan/mac-mcp (96 stars, last pushed yesterday), licensed MIT. It adds 87 tokens to every session and 819 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-21.
Other skills, from other repositories
b123d-modeling
Use this skill when asked to model, build, or modify a 3D part or assembly with build123d — from a text description, a technical drawing (image or PDF), dimensions in a spec, or an existing STEP/STL file.
b123d-edit
Use this skill when the task is to change an existing build123d model or script: "make this bracket 5 mm taller", "move the holes inward", "add a counterbore", "remove this rib", "make it fit this mating part", or "repair the code so the intended edit remains parametric".
authoring-decorators
Use when the user wants to write a new prompt decorator - either a personal one for their repo or a contribution back to the upstream catalogue. Walks through the JSON schema, the extensions/ convention, how to verify a new decorator end-to-end through the hook, and how to diagnose failures using debug logging and…
hypruse
Sees and drives the live Hyprland (Wayland) desktop through the hypruse CLI: lists windows and workspaces, focuses, moves, closes and launches apps over IPC, takes screenshots and zooms, reads GTK/Qt controls by name from the accessibility tree and clicks them, types text, presses shortcuts, runs the owner's keybinds…
video-to-skill
Author an ordinary host skill from retained video or demonstration observations, then validate its source provenance, timestamped evidence, assets, rights declarations and retained task-run bindings before producing a local deterministic .skill archive. Use when deriving reusable instructions from a recording; task…
gemini-visualize
Generates interactive HTML visualizations (concept maps, evidence networks, knowledge graphs) from Gemini analysis results. Use when the user requests a concept map, evidence network, or knowledge graph from completed analysis.