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 TryWorld2026/paper-algorithm --skill tryworld-kaiyangit clone --depth 1 https://github.com/TryWorld2026/paper-algorithmWrote 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/tryworld2026/paper-algorithm/tryworld-kaiyan)<a href="https://agentmods.dev/skills/tryworld2026/paper-algorithm/tryworld-kaiyan"><img src="https://agentmods.dev/badge/skills/tryworld2026/paper-algorithm/tryworld-kaiyan/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/tryworld2026/paper-algorithm/tryworld-kaiyan"><img src="https://agentmods.dev/badge/skills/tryworld2026/paper-algorithm/tryworld-kaiyan.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 | $0.00129 | $0.05033 |
| Opus 5.5 | $0.00052 | $0.02013 |
| Sonnet 5.5 | $0.00026 | $0.01007 |
| Haiku 4.5 | $0.00013 | $0.00503 |
Grade A, and why
tryworld-kaiyan 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 today.
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 — 343 lines — stays where its author put it; the contents beside it link to each section on GitHub.
试界开言
把选题或素材变成一份能直接开口讲的中文短视频稿。默认最终画面是真人主理人、主画面信息和真人小窗,但这里只写文字,不做视频。
先判模式
每次只进入一种模式。
- 选题模式:用户没有给明确主题,只说要口播选题。给出 3 到 5 张选题卡后停下,等用户选择。
- 成稿模式:用户给了主题、链接、新闻、笔记、旧稿或明确角度。核对材料后直接写出完整写稿包。
- 改稿模式:用户给了现成口播稿。保留用户确认过的事实和亲身经历,只改说话位置、结构、节奏、证据边界和平台适配。
用户要成片、配音、封面或剪辑时,说明本 Skill 只交付口播文字,可另行使用现有 tryworld-koubo 或 tryworld-paper。不要把这两个 Skill 变成运行依赖,也不要在这里启动 Setpiece、渲染或发布流程。
产品约束
- 受众是普通人。具体指非技术岗位的上班族、大学生、刚进入社会的年轻人,以及接触过 AI 但不是从业者、不是大佬、不是行业精英的人。他们不追 AI 新闻,也不写代码,但听得懂具体场景和好处。
- 长期选题约 60% 讲 AI 与普通人的机会,40% 讲 AI 工具和效率。单条视频只讲一个判断,不要求每条同时覆盖两类。
- 说话人是试界 TryWorld 主理人。第一人称,有判断,有取舍,不装专家,不喊口号,不假装亲历。
- 默认主稿 550 到 800 个汉字,按每分钟 260 个汉字估算。60 秒短版 300 到 450 个汉字。
- 默认五段推进,钩子、利害关系、两到三个证据或案例、边界或反常识、行动或收尾。自然段可以多于五段。
- 主稿默认用不超过 2 秒的短品牌签名开头,
哈喽,我是TryWorld。之后立刻给冲突、代价或反常识结论。抖音和小红书默认冷开场,把品牌识别放到结尾。 - 不建议使用“哈喽大家好,欢迎来到我的频道,我是TryWorld,今天我们聊聊……”这类长版频道开场。它会把真正的钩子推到 5 秒以后;用户坚持时先说明代价,标成人工例外,默认交付仍改成短签名。
- 每个自然段都要新增事实、动作、例子、区别或后果。只换说法、不加信息的段落删掉。
- 术语有预算。60 秒最多 1 到 2 个绕不开的词,第一次出现必须用普通话解释;能用具体动作和场景说清的,不用术语。
- 所有数字、日期、价格、产品能力、身份和引语必须有来源。来源不能只停在 AIHOT 摘要,重要事实回到原文。
- 用户没提供真实体验时,不写“我试过”“我用了一周”“我朋友亲测”。改用“官方说明”“公开资料显示”。
- 结尾先给当期内容需要的具体判断或行动,再用固定品牌署名收束。默认写
我是TryWorld,下次见。,也可用更亲切的我是TryWorld,下次见,拜拜。,不加问号。固定署名不能替代内容结论,也不强行点赞关注。 - 标题是交付物的一部分。主标题不超过 8 个字符,含标点,空格不计;目标 6 到 7 个字。抖音和视频号再压到 6 个字以内,小红书和 B站不超过 8 个字。
详细规则按需读取。
- 人格、受众和文字边界读
references/positioning.md - 选题卡、优先级和去重读
references/topics.md - 结构、口播句和七遍改稿读
references/writing.md - 来源、引语和事实边界读
references/evidence.md - 抖音、视频号、小红书、B站变体读
references/platforms.md - 质检闸门和对抗式审查读
references/quality.md - 标题规则、字数计算和三轮评审读
references/titles.md
工作流
1. 选题模式
先读 state/topic-history.json,避免重复最近已经写过的选题、角度和判断。
运行:
python scripts/fetch_topics.py --days 7 --take 100
脚本默认带 gzip、ETag/304 缓存;需要持续轮询时加 --watch,--interval 不得小于 60 秒。正常选题只取一次,不需要开 --watch。
抓取失败时不要卡住。明确告诉用户 AIHOT 暂时不可用,并请用户直接给一个主题、链接或素材。不要自己编造“最近热门”。
AIHOT 只负责发现线索。允许结合网络搜索、官方公告、产品文档、价格页和用户提供的素材。不能只按 AIHOT 热度排序,也不能把 AIHOT 摘要当成唯一事实来源。
选题先过普通人相关性硬闸门。每张卡必须回答五个问题。
What ships with it
20 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.
- .gitignore 66 B
- examples/opportunity-ai-actions.md 14 KB
- examples/tool-claude-meetings.md 14 KB
- LICENSES/human-writing-MIT.txt 1.4 KB
- README.md 3.9 KB
- references/evidence.md 8.1 KB
- references/platforms.md 7.6 KB
- references/positioning.md 7.3 KB
- references/quality.md 11 KB
- references/titles.md 9.6 KB
- references/topics.md 8.5 KB
- references/writing.md 12 KB
- scripts/check_script.py 29 KB runs code
- scripts/fetch_topics.py 16 KB runs code
- state/topic-history.json 19 B
- templates/topic-card.md 2.5 KB
- templates/writing-pack.md 9.8 KB
- tests/test_check_script.py 22 KB runs code
- tests/test_fetch_topics.py 14 KB runs code
- tests/test_writing_pack.py 12 KB runs code
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.
- today First seen · 343 lines · 129 tokens per session scan A 2a9051b04fb4
tryworld-kaiyan is a skill published in the GitHub repository TryWorld2026/paper-algorithm (34 stars, last pushed yesterday), licensed MIT. It adds 129 tokens to every session and 5,033 once invoked, about $0.0005 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 2026-11-16.
tryworld-kaiyan compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| tryworld-kaiyanTryWorld2026/paper-algorithm | Skill | 34 | yesterday | MIT | A | 129 | Claude Code, Codex |
| agent-wikiDianel555/DSkills | Skill | 65 | 13d ago | MIT | A | 89 | Claude Code, Codex |
| ace-toolDianel555/DSkills | Skill | 65 | 13d ago | MIT | A | 124 | Claude Code, Codex |
| github-trending-analyzerDianel555/DSkills | Skill | 65 | 13d ago | MIT | A | 65 | Claude Code, Codex |
| ledger-tasks-yyloDianel555/DSkills | Skill | 65 | 13d ago | MIT | A | 49 | Claude Code |
| capability-evolverDianel555/DSkills | Skill | 65 | 13d ago | MIT | A | 98 | Claude Code |
| cc-agyDianel555/DSkills | Skill | 65 | 13d ago | MIT | A | 187 | Claude Code, Codex |