Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/manaflow-ai/cmux-skillsnpx agentmods add skills/manaflow-ai/cmux-skills/cmux-workspaceWrote 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/manaflow-ai/cmux-skills/cmux-workspace)<a href="https://agentmods.dev/skills/manaflow-ai/cmux-skills/cmux-workspace"><img src="https://agentmods.dev/badge/skills/manaflow-ai/cmux-skills/cmux-workspace/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/manaflow-ai/cmux-skills/cmux-workspace"><img src="https://agentmods.dev/badge/skills/manaflow-ai/cmux-skills/cmux-workspace.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- Socket pass
- Snyk pass
- NVIDIA SkillSpector pass
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 | $0.00047 | $0.03492 |
| Opus 5 | $0.00023 | $0.01746 |
| Sonnet 5 | $0.00009 | $0.00698 |
| Haiku 4.5 | $0.00005 | $0.00349 |
Grade A, and why
cmux-workspace scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -fsS "http://127.0.0.1:<printed-port>" How it starts
The opening of the file, as written. The whole thing — 297 lines — stays where its author put it; the contents beside it link to each section on GitHub.
cmux Workspace
Use this skill when a task should be scoped to the cmux workspace that invoked the agent. A workspace is the sidebar tab-like unit in cmux. It contains split panes, and each pane contains one or more surfaces. A surface is the terminal or browser session the user interacts with.
Default Rule
Scope actions to the current caller workspace unless the user explicitly asks for another workspace, another window, or global state.
Do not assume the visually focused cmux workspace is the right target. An agent can be running in one workspace while the user is looking at another. Prefer the caller environment first:
printf 'workspace=%s\nsurface=%s\nsocket=%s\n' \
"${CMUX_WORKSPACE_ID:-}" \
"${CMUX_SURFACE_ID:-}" \
"${CMUX_SOCKET_PATH:-}"
cmux identify --json
Use CMUX_WORKSPACE_ID as the default workspace anchor and CMUX_SURFACE_ID as the default caller terminal/surface anchor. If those are missing, use cmux identify --json and be explicit that you are using the currently focused cmux context.
Non-Disruptive Automation
The user may be visually focused on a different workspace, window, or app while an agent works in the caller workspace. Treat layout and focus as separate concerns. Never call focus-changing verbs speculatively.
Never call these without an explicit user ask:
select-workspaceswitches the visible sidebar tab.focus-pane/focus-panelyanks pane or surface focus.tab-actionwith focus-changing actions.
These are user-affecting actions, like clicks. The rule applies even inside the caller's own workspace, since the user may be looking elsewhere.
Build layout additively, in one shot. Prefer commands that create a new pane already populated with the right surface:
# pane and content in one call, no follow-up needed
cmux new-pane --workspace "${CMUX_WORKSPACE_ID}" --type browser --direction right --url "http://127.0.0.1:8765"
cmux new-pane --workspace "${CMUX_WORKSPACE_ID}" --type terminal --direction down
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 12d ago First seen · 297 lines · 47 tokens per session scan A bf070d8d77de
cmux-workspace is a skill published in the GitHub repository manaflow-ai/cmux-skills (31 stars, last pushed 3d ago), licensed MIT. It adds 47 tokens to every session and 3,492 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
orbit-notion
Open Orbit briefing skill — selected by the Orbit pipeline when Notion is the user's only connected connector, or when the user explicitly scopes their daily digest to Notion. Pulls the past 24 hours of document edits, comments, mentions, and database row changes from the user's authenticated Notion connection and…
Cortex
Operate Cortex, the LifeOS memory system — the typed Knowledge Archive (People, Companies, Ideas, Research with typed related: links) plus recall of prior work sessions, ISAs, and conversations. Search, add, harvest, develop, ingest, distill, graph-navigate, recall. USE WHEN cortex, knowledge, knowledge base, search…
pinchtab-mcp
Use this skill when a task requires browser automation through PinchTab's MCP server connected to a remote browser instance. Covers navigation, element interaction, data extraction, form filling, multi-step flows, and session management via MCP tools.
feishu
Work with Feishu or Lark bots, docs, sheets, bitables, approval flows, and OpenAPI/MCP setup without hardcoding credentials.
peekaboo
Capture and automate macOS UI with the Peekaboo CLI.
mochi-remind
Handle due reminders — notify the user with natural language and mark them done.