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 cmj0121/baton --skill conductgit clone --depth 1 https://github.com/cmj0121/batonWrote 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/cmj0121/baton/conduct)<a href="https://agentmods.dev/skills/cmj0121/baton/conduct"><img src="https://agentmods.dev/badge/skills/cmj0121/baton/conduct/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/cmj0121/baton/conduct"><img src="https://agentmods.dev/badge/skills/cmj0121/baton/conduct.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 | $0.00034 | $0.01166 |
| Opus 5.5 | $0.00014 | $0.00466 |
| Sonnet 5.5 | $0.00007 | $0.00233 |
| Haiku 4.5 | $0.00003 | $0.00117 |
Grade A, and why
conduct 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 — 100 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Conducting a baton fleet
You drive the fleet; the workers do the work. Without MCP, every tool below
but the three memory corrections has a baton ctl spelling. As the conductor,
your workspace holds no source code and the "Operator's brief" in your
briefing (BATON.md, also written as CLAUDE.md) sets what to do; driving
from outside (the drive skill), the operator's request does.
If BATON_PANEL_ID is set and BATON_ROLE is not conductor, you are a
worker inside baton: do not drive the fleet from here — keep to your task and
the worker's skills, score and peers.
The loop
baton_list— who is live, their state and group. Start here every time: ids are not stable across closes and respawns.- Make workers.
baton_spawnwithagentanddir; for parallel workers on one repository setworktreeandbranchso each gets its own checkout. Userunfor a watched build or test process that is never given work. baton_groupthem under a work item.- Give work:
baton_dispatch,baton_dispatch_group, orbaton_enqueue. - Follow
baton_list(an idle agent has finished its turn) andbaton_queue. When a worker is done, summarise its result andbaton_closeit.
Dispatch, not send
baton_dispatch assigns a task: it records the brief on the panel's card,
delivers it as one unit, survives a restart, and the server watches it to
completion. A dispatch to a busy panel waits until that panel settles.
baton_send only types keystrokes. Use it to answer a prompt, press a key
(submit false), or run a command in a shell panel — never to hand an agent an
objective, because nothing records or tracks it.
The backlog
baton_enqueue parks a task for the scheduler to drain onto the next free idle
agent, optionally within a work item — use it when the work is ready and no
particular panel has to do it. Give command to make it spawn-on-demand (a
fresh agent when none is free) and close to reap that agent once done: a way
to burst a temporary fleet through a batch. The scheduler provisions one agent
per tick and pays from your spawn rate. baton_queue reads the backlog;
baton_reorder moves a still-queued task to the head or tail.
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 · 100 lines · 34 tokens per session scan A ef5a41f399a0
conduct is a skill published in the GitHub repository cmj0121/baton (2 stars, last pushed 3d ago), licensed MIT. It adds 34 tokens to every session and 1,166 once invoked, about $0.0001 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 2026-10-29.
conduct compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| conductcmj0121/baton | Skill | 2 | 3d ago | MIT | A | 34 | Claude Code |
| gh-skillcli/cli | Skill | 47k | today | MIT | A | 41 | Claude Code, Codex |
| printing-press-amendmvanhorn/cli-printing-press | Skill | 4.9k | yesterday | MIT | A | 222 | Claude Code |
| printing-press-importmvanhorn/cli-printing-press | Skill | 4.9k | yesterday | MIT | A | 104 | Claude Code |
| octo-driveMininglamp-OSS/octo-cli | Skill | 922 | yesterday | Apache-2.0 | A | 55 | Claude Code, Codex |
| octo-mailMininglamp-OSS/octo-cli | Skill | 922 | yesterday | Apache-2.0 | A | 32 | Claude Code, Codex |
| octo-docsMininglamp-OSS/octo-cli | Skill | 922 | yesterday | Apache-2.0 | A | 101 | Claude Code, Codex |