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 phuongduyphan/cc-tmux-agents --skill claudegit clone --depth 1 https://github.com/phuongduyphan/cc-tmux-agentsWrote 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/phuongduyphan/cc-tmux-agents/claude)<a href="https://agentmods.dev/skills/phuongduyphan/cc-tmux-agents/claude"><img src="https://agentmods.dev/badge/skills/phuongduyphan/cc-tmux-agents/claude/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/phuongduyphan/cc-tmux-agents/claude"><img src="https://agentmods.dev/badge/skills/phuongduyphan/cc-tmux-agents/claude.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 | $0.00039 | $0.01437 |
| Opus 5 | $0.00019 | $0.00718 |
| Sonnet 5 | $0.00008 | $0.00287 |
| Haiku 4.5 | $0.00004 | $0.00144 |
Grade A, and why
claude 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 11d 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 — 80 lines — stays where its author put it; the contents beside it link to each section on GitHub.
$claude — delegate to Claude Code CLI
Spawns Claude Code in a detached tmux session so Codex can offload work to it. The user stays in the loop: they ask Codex when to check progress, what guidance to send, and when to stop.
Primitive
All operations go through a single dispatcher:
~/.agents/skills/claude/bin/claude-agent [--session <name>] <verb> [args]
Verbs:
start [task...]— create session (defaultcodex-claude), launchclaude --dangerously-skip-permissions, type the task and press Enter. Returns immediately (non-blocking).send <message>— type message + Enter. For follow-up guidance. Does not interrupt current generation; Claude queues the input.keys <name>— send a raw tmux key (Escape,C-c,Enter, etc.). Used for interrupts.capture [lines]— print last N ANSI-stripped rendered lines (default 500). Read-only, non-destructive.wait [stable] [timeout]— block until the agent goes idle, then exit. Run it in the foreground to block until the sub-agent finishes (the mode for Codex and other CLI orchestrators; see "Wait for completion" below).stable= secs the working indicator must stay gone before declaring idle (default 6);timeout= overall cap in secs (default 1800).stop— kill-session. Destroys the tmux session.list— list activecodex-claude*sessions.
How to respond to the user
| User says | Do |
|---|---|
$claude <task> / "delegate to claude" |
start <task>, then run wait in the foreground (blocks this turn until claude is idle); on return, capture + report. |
$claude <task> + "tell me when it's done" / "notify me" |
Same as $claude <task>; foreground wait is already the default. |
| "check claude" / "progress" / "what's claude doing" / "ask claude for progress" | capture. Show the relevant tail, add a one-line diagnosis (running / idle / error / prompt). Do not send ESC. |
| "interrupt claude" / "esc claude" | keys Escape → capture. |
| "tell claude: X" / "send claude: X" | send X. |
| "kill claude" / "stop claude" | stop. |
What ships with it
1 file 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.
- 11d ago First seen · 80 lines · 39 tokens per session scan A 9dd5fbd4cab3
claude is a skill published in the GitHub repository phuongduyphan/cc-tmux-agents (8 stars, last pushed 1mo ago), licensed MIT. It adds 39 tokens to every session and 1,437 once invoked, about $0.0002 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
repo-hygiene
Use when the scheduled repo-hygiene workflow runs from GitHub Actions (or an operator dry-run) to scan the repository for small, certain docs/test/code hygiene issues and fix them as one batched branch.
find-simplifications
Use for a periodic repo-wide sweep of qwen-code for accumulated excess surface — dead components and files, orphaned locale keys, exports nothing consumes, added-then-removed scaffolding — filing candidates on a tracking issue and landing only what a maintainer has said yes to. Repo-wide and evidence-first; every…
autoprompt
Run explicitly requested Autoprompt work with task routing, owned assignments, independent checks, and bounded recovery.
autoprompt
Run explicitly requested Autoprompt v2 work through the private controller. Ordinary coding and review requests do not activate this skill.
ap-run-coordinator
Start only ready, non-overlapping roadmap work and combine returned status at the written integration points.
performance-optimization
Measure-first performance work. Use on triggers like "it's slow", "profile this", "optimize perf", "fix the bottleneck", "improve load time / Core Web Vitals", or when a measured regression needs fixing. Enforces measure-before-optimize. Fills a perf gap not covered by existing project skills. Not a build/ship…