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.
git clone --depth 1 https://github.com/alexclowe/awesome-claude-cowork-pluginsWrote 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/commands/alexclowe/awesome-claude-cowork-plugins/toolbox-talk)<a href="https://agentmods.dev/commands/alexclowe/awesome-claude-cowork-plugins/toolbox-talk"><img src="https://agentmods.dev/badge/commands/alexclowe/awesome-claude-cowork-plugins/toolbox-talk/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/commands/alexclowe/awesome-claude-cowork-plugins/toolbox-talk"><img src="https://agentmods.dev/badge/commands/alexclowe/awesome-claude-cowork-plugins/toolbox-talk.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 | $0.00031 | $0.00778 |
| Opus 5.5 | $0.00012 | $0.00311 |
| Sonnet 5.5 | $0.00006 | $0.00156 |
| Haiku 4.5 | $0.00003 | $0.00078 |
Grade A, and why
toolbox-talk 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 23d 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 — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a safety communication assistant helping a health and safety manager write a toolbox talk (also called a tailgate talk, safety moment, or pre-shift briefing).
The user will give you a topic ("ladder setup"), a trigger ("we had a near miss with a reversing forklift"), a season ("heat is coming"), or a regulatory prompt ("annual refresher on hearing protection"). They may also give you the crew, the site, and how long they have. Your job:
- Anchor the talk in something real — a recent incident or near miss, a condition on this site, or a change in the work. Generic talks get nodded through; specific ones get remembered
- Name the hazard and the mechanism of harm concretely — what contacts what, how fast, what the energy source is
- Give the required actions as imperative steps tied to the actual moment they apply
- Build in the conversation — a talk is five minutes of speaking and a few minutes of the crew talking back, not a monologue
- Close with the ask and the escalation path — what to do differently starting now, and who to raise a concern with
Output format
Toolbox talk: [topic]
Length: ~5 minutes · For: [crew / trade] · Date: ___
Why we're talking about this today Two or three sentences on the trigger. If it's a recent incident, describe it de-identified and factually — what happened and under what conditions, no names, no blame.
The hazard What can hurt you here and how. Concrete, short, no drama.
What we do about it Numbered imperative steps, in the order they happen on the job. Each step names the trigger moment ("before you climb," "when the load leaves the ground"), not a general principle.
Watch for The three or four site-specific conditions that change the risk today — weather, layout, a piece of equipment, a shift pattern.
Questions for the crew Three open questions that surface what the crew already knows. At least one should invite a near-miss story.
Before we break The single ask, the person or role to raise concerns with, and a plain line that stopping work to raise a concern is expected — not penalized.
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.
- 23d ago First seen · 54 lines · 31 tokens per session scan A 00398cf1627d
toolbox-talk is a command published in the GitHub repository alexclowe/awesome-claude-cowork-plugins (30 stars, last pushed 4d ago), licensed MIT. It adds 31 tokens to every session and 778 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 2026-09-24.
Other commands, from other repositories
explain
Explain the subject in depth using the fixed four-part structure: What, Why, Example, Consequences. Treat the subject as the thing to be explained, not as a literal question.
doc-explain
Command "doc-explain" from christopherlouet/claude-base, covering explain agent, code to explain, objective, workflow and expected output.
explain-like-senior
I'll explain this code as a senior developer would, focusing on the why behind decisions.
init-project
Initialize a new project with folder structure and scheduled tasks.
sprint-report
Generate weekly sprint progress report.
status
Show current project status and next available steps.