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 pranav7/domux --skill domux-communicategit clone --depth 1 https://github.com/pranav7/domuxWrote 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/pranav7/domux/domux-communicate)<a href="https://agentmods.dev/skills/pranav7/domux/domux-communicate"><img src="https://agentmods.dev/badge/skills/pranav7/domux/domux-communicate/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/pranav7/domux/domux-communicate"><img src="https://agentmods.dev/badge/skills/pranav7/domux/domux-communicate.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 | $0.00089 | $0.00936 |
| Opus 5 | $0.00044 | $0.00468 |
| Sonnet 5 | $0.00018 | $0.00187 |
| Haiku 4.5 | $0.00009 | $0.00094 |
Grade A, and why
domux-communicate 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 9d 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 — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/domux-communicate — message a peer agent in another worktree
domux runs each worktree's Claude agent in its own tmux pane. These three commands let you address a peer agent by worktree name — you never scan panes yourself; domux resolves the name to the right pane from the session mapping it already maintains.
The canonical use is a handoff: you finished something on your branch and the agent in another worktree needs it ("I built X on branch Y, here's the doc, please pull it in").
The peer is another Claude agent, not the human operator. Every message you send is automatically prefixed to say so, so the peer doesn't mistake it for its user.
1. Discover who's running — domux peek
domux peek
Lists every running Claude agent across worktrees: its NAME (what you pass
to send/read), STATE (working / waiting / idle), TASK, and tmux
TARGET. Run this first if you're unsure of the exact name.
2. Send a message — domux send <name> <message…>
domux send workspace-2 "I pushed the fix on branch eng-225-agent-calc. Pull it: git fetch && git checkout eng-225-agent-calc. The doc is docs/eng-225.md."
- Name is the worktree's session name, its directory/branch basename, or its domux label — any of them resolve.
- Flags come before the positionals (Go flag parsing stops at the first
non-flag word):
--from NAME— how to attribute the message. Defaults to your own session's label or name, so the peer sees who it's from.--pane W.P— pick a specific pane (e.g.--pane 2.0) when a session runs more than one agent.--no-enter— type the message into the peer's input box but don't submit it, so a human (or you, later) can review before sending.--wait(with optional--wait-timeout 2m) — block until the peer is idle, then send. Without it, the message sends immediately; if the peer is mid-generation Claude Code queues it (nothing is lost) andsendtells you it was queued.
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.
- 9d ago First seen · 82 lines · 89 tokens per session scan A 7eb74a8fd2cb
domux-communicate is a skill published in the GitHub repository pranav7/domux (5 stars, last pushed today), licensed MIT. It adds 89 tokens to every session and 936 once invoked, about $0.0004 per session on Opus 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 2026-08-31.
Other skills, from other repositories
crawlkit
Maintain and release the crawlkit Go library, preserving downstream compatibility for gitcrawl, slacrawl, discrawl, and notcrawl.
github-issue-fix-flow
Use when asked to take a GitHub issue number, implement a fix, run builds/tests, commit with a closing message, and push. End-to-end workflow using gh CLI, local code changes, and git.
release-app-store-changelog
Create user-facing App Store release notes from git history. Use when asked to generate release changelog, App Store "What's New" text, or release notes based on git tags.
manager-loop
Run a large batch of independent work items (a fleet-wide harvest, a multi-repo sync-down, an overnight backlog) as a persistent coordinator loop instead of one item at a time. Use when there are enough independent units of work that dispatch-review-merge would otherwise repeat many times in a row, or when the batch…
goal-orchestrator
Run a long-lived, headless per-project GOAL orchestrator on hyperpanes — hold a project's goal list, spawn a fable/opus spec agent per goal, have it fan work out to sonnet impl agents via the durable work queue, watchdog wedged agents, rotate across Claude accounts on limits, and loop 24/7. Use when the user wants a…
release
Build, version bump, update README/marketplace, commit, push, create GitHub release, and clean local caches.