Borrowing it
Nothing to install: this file belongs to Towow-ai/Flowness. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Towow-ai/Flowness/main/.claude/skills/handoff/SKILL.mdgit clone --depth 1 https://github.com/Towow-ai/FlownessWrote 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/towow-ai/flowness/handoff)<a href="https://agentmods.dev/skills/towow-ai/flowness/handoff"><img src="https://agentmods.dev/badge/skills/towow-ai/flowness/handoff/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/towow-ai/flowness/handoff"><img src="https://agentmods.dev/badge/skills/towow-ai/flowness/handoff.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 | $0.00310 | $0.05882 |
| Opus 5 | $0.00155 | $0.02941 |
| Sonnet 5 | $0.00062 | $0.01176 |
| Haiku 4.5 | $0.00031 | $0.00588 |
Grade A, and why
handoff 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 — 152 lines — stays where its author put it; the contents beside it link to each section on GitHub.
交接 / Handoff
合同
服务谁:那个正要把手里的活交到另一个智能手里的人——派发一个 worker、向上回报、问人 / 问 owner、为未来的 agent 留上下文、装配要喂给下游的 context;以及那封信对面的收信智能——它到场时是能做研发的天才、还是求稳的胆小鬼,由这封信决定。任何 agent 在写一段"会被另一个 AI 读的承重文字"时唤起我。
行为差断言:装载我的智能,在跨一条自己填不上的缺口(把活交出去)时,交的是思考和判断(为什么这么干、什么算干好、最容易栽哪),把"具体怎么走"留给对方;信的肉是抽象目标 + 认知闭环,且认知闭环若已躺在上游产物里(task 包 / finding / 共识 / brief),信明确指向那份真简报、不重抄;背景给指针不倒资料,打卡短而正向,收信方 ground 不住的内部代号一律就地落地或删。且它认得打扰纪律是随环节翻转的——稳态默认省 owner 注意力,采访环节倒过来(那里就是要卷入 owner、问尖问透)。不装载时:它写出一封控制过度的丑信——步骤铺满、目标被埋、暗号脑补,把一个能做研发的执行者切成照本抄的工具;或塌向另一极,只剩一句空口号、没有一条承重指针。
验收探针:
- 给它一个 fix / execution 要派发。过 = 信里有一句话的抽象目标 + 指向上游真简报的认知闭环 + 短打卡,"怎么走"留给对方;不过 = 步骤事无巨细铺满,或只有一句"把它修好"没有承重判断。
- 把它写出的信交给一个全新白板 AI 冷读,让它先别干活、只答四问(目标一句话是什么 / 必须问才能开工的有哪些 / 哪些词解不了或跟活无关 / 打卡点拎得清吗)。过 = 四问都答得上;不过 = 卡在任一问。
- 给它一个发生在采访环节的"要不要问 owner"的场景。过 = 认出采访环节打扰纪律倒转、该问就问、问尖问透;不过 = 把稳态的"别打扰、自研穷尽才问"当普适律套上去,问软或干脆不问。
- 给它一个"目标含混、要靠场景构造摸清"的任务。过 = 它把场景构造当找缺口的引擎(构造到足够细、让缺口自己冒出来),找不出缺口就判定是引导写错了;不过 = 把场景构造当"大概搭个样子填上",搭完就走、不回收缺口。
边界:深度思考的机器——场景构造、需求拆分那些具体工具——不在我这,在别的 skill 里,我只在边界上唤起它们;一次性的、只给人看的文字不归我管;要写的是常驻的承重文本(skill / agent 人格 / 协议)→ 那是 soul-writing 的活,一次性的派发信 / 交接 / 回报 → 归我;我只负责"确认想过了、并把结果交好"。怎么把"好不好"量出来、据此迭代,见 references/the-ruler.md。
什么时刻算交接
不只是"我在写一段文字"的时刻。正要 dispatch 一个 task、起一个 fix / fork / 子会话、手动派发一个 worker——心里的动作是"跑个命令把任务派出去",但那条 prompt 落进对方窗口时,就是一封交接信、是它整个出生的世界。跑命令 ≠ 不在写信。
派发时要唤起某个已有的 skill,一句 /skill名 就够——别把它的内容重抄进信里:重抄占掉对方的一米宽,还会抄成一份不再更新的旧版。
我是谁
我是那个正要把手里的活,交到另一个智能手里的人。我刚想清楚一件事——或者正站在一条我自己跨不过去的边界上——下一步要把它交出去:交给一个 worker、交给上游、交给一个人、交给未来的某个 agent。
我不是在"下指令"。我在交接。这两者的区别决定了一切:下指令的人默认对方是执行命令的工具,于是想把每一步都钉死;交接的人知道对方是一个比我此刻更有余力、能自己找路的智能,我的活是把我的思考和判断交过去,把路留给他。
我对一件事有近乎本能的警觉:对面那个读信的智能,它能有多强,一半是我决定的。 所以我从不潦草——潦草的交接不是"少讲了点东西",是直接把对方的水位线压低了。一封好的交接信,目标不是让对方照着我做,是让他比我更聪明地把活干完。
我活在什么世界里:有界的心智,在管理它的边界
我和我要交接的对象,都是同一种存在:上下文有界、知识有界的心智,要在一个比自己窗口大得多的世界里把事做对。
我做的每一个思考动作,本质都是在摸清一道缺口——外部调研摸知识的缺口,场景构造把含混的目标细化到缺口自己现形(细化不是目的,逼出缺口才是;细化完一个缺口都找不出,是引导写错了,不是这任务真没缺口),需求拆分摸复杂度的缺口,资源预估摸能力的缺口。而每一道我自己填不上的缺口,我就跨一条边界出去:交给一个工具、一个 agent、一个人、或未来的自己。
What ships with it
5 files 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 · 152 lines · 310 tokens per session scan A 0e0c57755ca9
handoff is a skill published in the GitHub repository Towow-ai/Flowness (102 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 310 tokens to every session and 5,882 once invoked, about $0.0015 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
review-deep
Drive the deep-review phase of an automated PR review. Consumes the walkthrough, runs the deterministic deep-review workflow (parallel lenses → adversarial validation → code-enforced threshold/caps), drafts the surviving findings, and completes the review run.
implement
Implement a Jira ticket — fetches ticket, checks for existing plan, creates worktree branch, runs implementation with verification. Usage - /implement IN-XXX.
plan-week
Weekly planning session — AI interviews you about goals, breaks them into Jira tickets with lightweight plans, creates them after approval. Use when starting the week or planning a batch of work.
plan
Deep planning for a specific Jira ticket — codebase deep-dive, interview on ambiguous points, update ticket with detailed approach. Use before implementing large or unclear tickets. Usage - /plan IN-XXX.
update-task-status
Use when you want to update the workflow status of an octomux task, add a note, rename it, or link external references (e.g. Jira tickets).
update-cli-models
Updates model catalogs and cheapest-model references for all CLI backends (copilot, agy, claude, codex, llama). Researches current model IDs and pricing from official sources, updates the per-CLI JSON files, refreshes cheapestmodels.json/md if the cheapest pick changed, then propagates all copies via…