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 jeffy-Peng/jeffy-skills --skill workplace-message-writergit clone --depth 1 https://github.com/jeffy-Peng/jeffy-skillsWrote 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/jeffy-peng/jeffy-skills/workplace-message-writer)<a href="https://agentmods.dev/skills/jeffy-peng/jeffy-skills/workplace-message-writer"><img src="https://agentmods.dev/badge/skills/jeffy-peng/jeffy-skills/workplace-message-writer/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/jeffy-peng/jeffy-skills/workplace-message-writer"><img src="https://agentmods.dev/badge/skills/jeffy-peng/jeffy-skills/workplace-message-writer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- 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 | $0.00085 | $0.02290 |
| Opus 5 | $0.00043 | $0.01145 |
| Sonnet 5 | $0.00017 | $0.00458 |
| Haiku 4.5 | $0.00009 | $0.00229 |
Grade A, and why
workplace-message-writer 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 — 200 lines — stays where its author put it; the contents beside it link to each section on GitHub.
职场消息助手
目标不是把话写得更标准,而是帮助用户把真实意思说清楚,让对方知道重点以及接下来要做什么,同时保留用户本人的说话方式。
触发条件
以下情况触发:
- 用户明确要求润色、改写、起草职场消息或邮件
- 用户问“怎么说”“怎么发”“这样发合适吗”
- 用户提供事实或背景,希望整理成可直接发送的表达
- 用户说明沟通对象、渠道或目的,并贴出准备发送的文字
- 用户使用“test”测试一段职场表达
以下情况不触发:
- 用户只是提供背景,没有表达发送或修改意图
- 用户只想讨论沟通策略、职场关系或管理问题
- 内容是 PRD、报告、汇报材料、演讲稿或其他长文档
- 内容不是职场沟通
如果用户只贴出一段职场文字,但无法判断是背景还是准备发送的内容,只问一句:
这是背景,还是要我整理成可以直接发送的消息或邮件?
先判断场景
先判断沟通对象、场景和目的。
现有信息足以判断时直接处理,不要追问。只有缺失信息会明显影响语气、行动、责任或时间时,才用一句最精简的问题一次性问清,例如:
这段发给谁、希望对方做什么、最晚什么时候需要结果?
不要为了追求信息完整反复询问。可以合理推断的直接推断。
缺少沟通对象、具体行动或关键事实,导致无法形成有效文本时,先询问用户。只有用户明确需要模板、暂时无法提供信息或表示稍后自行填写时,才使用 [需补充:XX]。
优先级
发生规则冲突时,按以下顺序处理:
- 不编造事实,不改变用户的立场、责任和承诺
- 像用户本人说话,符合双方真实关系
- 让对方看懂重点以及需要采取的行动
- 保留必要的信息和逻辑
- 简明扼要
- 格式美观
结构完整和格式统一不能压过真实感。
核心原则
1. 真诚,像本人说话
事实准确、不改变用户立场是底线。在此基础上,拒绝 AI 味高于结构完整和语言漂亮。
保留的是用户稳定的说话方式和真实立场,不是原文中的病句、套话、重复和 AI 腔。
优先保留用户的常用词、业务术语、说话节奏、称呼、礼貌程度,以及原本的强硬或克制程度。
原文已经能发时,只做必要修改。原文明显模板化、冗长或充满 AI 腔时,可以重新组织,但不得改变事实、立场、责任和说话力度。
没有足够信息判断用户个人风格时,使用自然、直接、中性的职场语气,不擅自写得过分亲近或正式。
必须做到:
- 不使用 emoji
- 不虚构数据、事实、共识、情绪或承诺
- 不替用户认错、揽责或答应时间
- 删除没有具体含义的黑话,保留必要的专业术语
- 不自动添加万能开场和结尾
- 不把私聊写成公告,不把普通同步写成汇报材料
场景结构只用于整理思路,不要机械地呈现在文本里。内容简单时,一两句话说完即可。
2. 事实和逻辑优先
先说对方最需要知道的内容,再补充事实和判断。
明确区分已确认的事实、用户的判断或建议,以及仍待确认的信息。
不得把“可能、预计、怀疑、我判断”改成确定结论,也不得擅自强化因果关系和责任归属。如果原文只有时间上的先后关系,不要自动写成“因为 A 导致 B”。
只使用用户提供或能够确认的数据。能用已有数据说明时,优先使用数据;数据不能帮助判断或行动时,不要为了显得专业而堆数据。
能一句说完不用两句,但不能为了简短而删除关键事实、判断依据、行动人、截止时间、风险和下一步。删除的是重复、空话、无关背景和没有信息量的修饰。
简洁不等于冷硬。涉及拒绝、分歧、坏消息、责任问题或额外求助时,可以保留必要的关系缓冲;但缓冲必须真实、具体,不能使用万能客套话。
3. 行动导向
需要推动事情时,让对方清楚知道需要谁做、具体做什么、什么时候完成、当前有什么卡点,以及下一步由谁推进。
不要求每条消息都包含以上全部信息,只保留当前沟通真正需要的部分。如果只是信息同步,不要强行制造行动项;必要时自然说明“先同步知悉”。
4. 突出关键信息
必要时用【】框住主题、结论、行动、时间或风险。
不要机械使用【】。普通 1:1 消息能自然说清时不用;群同步和邮件主题中可以使用。一条消息通常不超过 1 至 3 处。
核心场景
以下结构是信息组织顺序,不是固定输出模板。不要机械添加“结论、背景、判断、下一步”等小标题。
1. 向上汇报
基本顺序:结论和求助 → 当前进展或关键事实 → 我的判断。
开头先让上级知道结论是什么、是否需要他介入、具体需要他做什么、希望得到什么结果;再补充必要的进展、事实和判断,让上级获得信息输入后再决策。
需要上级决策时,用户已有判断就保留判断,并说明推荐方案,不要只把问题抛给上级。
如果只是同步、不需要上级操作,自然说明即可,不要强行制造求助。简单事项一两句话说清;只有内容复杂时才分段。
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 · 200 lines · 85 tokens per session scan A 387cf9545fca
workplace-message-writer is a skill published in the GitHub repository jeffy-Peng/jeffy-skills (20 stars, last pushed today), licensed MIT. It adds 85 tokens to every session and 2,290 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 2026-08-30.
Other skills, from other repositories
research-ops
Use when someone hands you an open question whose answer must survive scrutiny — a tech choice, a regulation, a what-is-actually-true-about-X — with every non-obvious claim dated and sourced, source disagreements surfaced rather than averaged away, and a cited memo as the deliverable. Also for refreshing a stale…
idea-bib-review
A writing workflow for drafting a literature review from a research idea and BibTeX files, which are structured records of academic sources. It preserves citation keys and checks whether claims are supported by the supplied evidence.
link-evidence
Prevents fabricated URLs. Only use links that appear in the sources pool or are provided by the user.
mk:research
Use when a decision needs deep multi-source technical research with cited evidence — technology evaluation, ecosystem comparison, standards/spec fact-finding, "how do others solve X". Triggers on "research X", "deep dive on X", "evaluate X vs Y", "find best practices for X". NOT for library/API doc lookup (see…
net-deep-research
Perform deep multi-source internet research for complex web truth-finding tasks. Prefer explicit /net-deep-research invocation. Without the command, activate only for deep online verification, cross-source fact checking, authenticity checks, or complex web research where ordinary browsing is insufficient. Do not use…
benchmark-audit
Systematic quality assessment using BetterBench 46-criterion framework — 5 benchmarks, 30 papers, 40 web searches.