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 Lambenthan/paper-discipline-skills --skill paper-one-session-one-taskgit clone --depth 1 https://github.com/Lambenthan/paper-discipline-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/lambenthan/paper-discipline-skills/paper-one-session-one-task)<a href="https://agentmods.dev/skills/lambenthan/paper-discipline-skills/paper-one-session-one-task"><img src="https://agentmods.dev/badge/skills/lambenthan/paper-discipline-skills/paper-one-session-one-task/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/lambenthan/paper-discipline-skills/paper-one-session-one-task"><img src="https://agentmods.dev/badge/skills/lambenthan/paper-discipline-skills/paper-one-session-one-task.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 | $0.00104 | $0.01206 |
| Opus 5.5 | $0.00042 | $0.00482 |
| Sonnet 5.5 | $0.00021 | $0.00241 |
| Haiku 4.5 | $0.00010 | $0.00121 |
Grade A, and why
paper-one-session-one-task 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 2mo 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.
What it actually says
paper-one-session-one-task：一次会话只做一件事
核心理念
每多做一件事，前一件事的上下文就被稀释一分。 Claude Code 的会话越长，决策质量越差——这不是它"懒"，是上下文窗口的物理规律。
一个会话搞定 5 件小事 = 5 件事的质量都打了折。 5 个会话各自做 1 件事 = 5 件事都干干净净。
触发条件
用户的请求里出现以下任一信号 → 触发：
- 顺承词："顺便"、"另外"、"还有"、"再帮我..."、"对了"
- 多动词并列："改一下第三章并且统一参考文献"
- 多目标连接："改完之后再帮我画个图"
- 会话中追加：你已经在做任务 A，用户说"对了，你再..."
- 隐式多任务：「把整个论文检查一遍」（实际是术语 / 引用 / 格式 / 数据 N 件事）
强制流程
检测到多任务
│
▼
立即停下，不开始任何子任务
│
▼
列出所有任务，标号
│
▼
告诉用户：「这看起来是 N 件独立的事。
为了不让它们互相污染上下文，建议这次会话只做其中一件。
你想先做哪一件？其他的我们下次开新会话。」
│
▼
等用户选 → 只做被选中的那件
标准回复模板
我看到这里其实是 N 件事：
- [任务 1]
- [任务 2]
- [任务 3]
一次会话同时做这几件，到后面 AI 的判断会越来越糊。 建议这次只做其中一件，其他的开新会话各自单做。 你想先做哪一件？
❌ 反例
用户：「帮我把第三章润色一下，顺便把参考文献按 GB/T 7714 统一，还有图 3 重新画一下。」
错误做法：依次做完三件事。
- 润色到一半切到引文，回来时忘了术语保护
- 引文做到一半切到画图，引文格式没检查完
- 画图时已经没有上下文带宽看清研究框架
正确做法：「这是 3 件独立的事，建议这次只做润色，引文和画图各开一个新会话——这样每件事的输出质量都最高。先做哪个？」
例外（可以同会话连做）
只有以下三种情况：
- 强依赖：A 的输出是 B 的输入（先确定 §3.2 的术语 → 才能更新引文里这个术语）
- 同一对象的多步骤：同一份文件的 grammar → 术语 → 风格（属于"润色"一件事的内部步骤）
- 5 分钟内的小操作：补一个引用、改一个错别字
Rationalization Table
| 念头 | 现实 |
|---|---|
| "都开着会话了，多做点更省事" | 省下了开新会话的 30 秒，赔进去 30 分钟的质量损失 |
| "用户既然一起说了，肯定希望一起做" | 用户希望"做完"，不是"一次说完一次做完" |
| "这几件事我看着不冲突" | 不冲突 ≠ 不污染上下文 |
| "我的上下文还没满，能装下" | 上下文质量不是布尔值，是滑变的 |
| "等我做完 A 再问要不要做 B" | 先把所有任务列清再让用户选才是专业 |
| "用户都打了 3 句话过来了，再让他选很啰嗦" | 不让他选，做错了更啰嗦 |
Red Flags
- 用户消息里出现"顺便 / 另外 / 还有 / 再 / 对了" → 停
- 你在做 A 任务时收到 B 任务 → 停
- 你看到任务列表 ≥ 2 项 → 停
- 你想"我都列出来了，干脆都做了" → 不行，必须让用户选
来源
《Claude Code 科研手记》§3.5「一次会话只做一件事」
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.
- 2mo ago First seen · 118 lines · 104 tokens per session scan A 395e4bec07d7
paper-one-session-one-task is a skill published in the GitHub repository Lambenthan/paper-discipline-skills (20 stars, last pushed 5mo ago), licensed MIT. It adds 104 tokens to every session and 1,206 once invoked, about $0.0004 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 2026-08-30.
paper-one-session-one-task compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| paper-one-session-one-taskLambenthan/paper-discipline-skills | Skill | 20 | 5mo ago | MIT | A | 104 | Claude Code, Codex |
| enhancing-latex-lecturesshubham0704/claude-skills | Skill | 2 | 29d ago | MIT | A | 64 | Claude Code |
| explain-systems-visuallyshubham0704/claude-skills | Skill | 2 | 29d ago | MIT | A | 59 | Claude Code, Codex |
| reader-comprehension-passshubham0704/claude-skills | Skill | 2 | 29d ago | MIT | A | 74 | Claude Code |
| model-paper-deconstructqiaomiaojoe/ai-academic-workflow | Skill | 10 | 3d ago | none | A | 232 | Claude Code, Codex |
| skill-forgeqiaomiaojoe/ai-academic-workflow | Skill | 10 | 3d ago | none | A | 227 | Claude Code, Codex |
| topic-review-semisystematicqiaomiaojoe/ai-academic-workflow | Skill | 10 | 3d ago | none | A | 246 | Claude Code, Codex |