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 0xMulight/human-social-copy --skill ai-tips-tricksgit clone --depth 1 https://github.com/0xMulight/human-social-copyWrote 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/0xmulight/human-social-copy/ai-tips-tricks)<a href="https://agentmods.dev/skills/0xmulight/human-social-copy/ai-tips-tricks"><img src="https://agentmods.dev/badge/skills/0xmulight/human-social-copy/ai-tips-tricks/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/0xmulight/human-social-copy/ai-tips-tricks"><img src="https://agentmods.dev/badge/skills/0xmulight/human-social-copy/ai-tips-tricks.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 | $0.00053 | $0.05404 |
| Opus 5 | $0.00026 | $0.02702 |
| Sonnet 5 | $0.00011 | $0.01081 |
| Haiku 4.5 | $0.00005 | $0.00540 |
Grade C, and why
ai-tips-tricks scanned grade C with 2 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 12d 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.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
1. **curl → python3 必须分两步，不能 pipe。** `curl ... | python3 -c "..."` 触发 `tirith:curl_pipe_shell` 直接被拒。正确做法：`curl -s -o /tmp/gn.xml "..."` 然后 `python3 -c "..."` 读文件。每次搜索前先确认用的是两步法。 Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
1. **curl → python3 必须分两步，不能 pipe。** `curl ... | python3 -c "..."` 触发 `tirith:curl_pipe_shell` 直接被拒。正确做法：`curl -s -o /tmp/gn.xml "..."` 然后 `python3 -c "..."` 读文件。每次搜索前先确认用的是两步法。 How it starts
The opening of the file, as written. The whole thing — 328 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI Tips & Tricks Pipeline
全网搜索最新AI玩法/技巧/工作流，写成教程式中文推文推送。
内容类型标准（v1.4 新增）
🔴 每条推送必须是可复制的具体操作，不是泛泛观点。判断标准：读者看完能不能打开AI工具立刻复制使用？
✅ 推这类：
- Agent配置技巧（子代理设置、模型搭配、参数调优）：如 Codex luna-max 并行执行、sol+luna 分工
- Prompt模板（有完整可复制文本）
- 工作流组合（"先让A做X，再让B做Y，最后得到Z"，有具体步骤）
- 隐藏功能/反常识操作（大多数人不知道的用法）
- 免费替代付费的实操方案
❌ 不推：
- 纯工具推荐（没有技巧成分）
- GitHub 仓库
- "10个XX技巧"列表
- 产品功能公告/更新日志
- 需要编程/API Key 才能用的
🔴 CRON EXECUTION CHECKLIST（每次执行前必读）
- curl → python3 必须分两步，不能 pipe。
curl ... | python3 -c "..."触发tirith:curl_pipe_shell直接被拒。正确做法：curl -s -o /tmp/gn.xml "..."然后python3 -c "..."读文件。每次搜索前先确认用的是两步法。 - Valnet 属性全部跳过。 MakeUseOf、How-To Geek、Tom's Guide、XDA、Android Authority、Android Police 等 r.jina.ai 全部返回导航栏（3-8KB）。不要在这些域名上浪费提取尝试。
- 产品公告不选。 ChatGPT Voice 桌面版、ChatGPT Health 开放等产品上线新闻，中文媒体几小时内覆盖，无信息差窗口。
- 至少搜 4-6 组不同关键词。 轮换有效查询模式，不要重复同一组。
- 每个 Google News RSS 搜索结果 ≤ 10-15 条提取，避免信息过载。
搜索策略
⚠️ Cron job 环境中没有原生 web_search 工具，所有搜索通过 curl + terminal 完成。 详见 references/search-reliability.md 和 references/extraction-degradation-2026-07-27.md。
搜索优先级（按可靠性排序）
- Google News RSS（最可靠）— 始终返回结果，带日期和来源
- r.jina.ai — 用于提取文章正文。⚠️ 仅 Hackaday、Ars Technica、HackerNoon、Coursera、学术 .edu 域名（MIT Sloan等）、个人博客可提取。PCWorld/CNET/Yahoo/Medium/Valnet 全系/TechCrunch/businesstoday.in/Business Insider 均失效。
- HN Algolia API — 可靠但结果偏开发者向，偶尔有可用内容
以下方法不要用：DDG Lite（严重限速）、Reddit JSON API（返回HTML）、Google curl（空结果）、Bing curl（只返回官方页面）、Google Cache（重定向）。
查询策略
不要直接搜 tip/trick/hack，这只会返回泛泛的列表文章。改用结果导向的查询：
有效查询模式（实测可产出个人体验/实操类内容）：
"how I use" OR "changed my" OR "game changer" ChatGPT OR Claude 2026← 个人体验角度，最有效"most people ignore" OR "underrated" OR "hidden" ChatGPT OR Claude feature 2026← 隐藏功能"saves me" OR "changed how" OR "every day" ChatGPT OR Claude 2026← 效率提升ChatGPT image generation prompts create edit 2026← 特定场景"prompt that" OR "try this prompt" OR "copy this prompt" ChatGPT OR Claude 2026← 寻找可复制的 prompt 模板"backwards" OR "reverse" OR "invert" ChatGPT OR Claude prompt technique 2026← 特定技巧名搜索，产出的都是教程类而非新闻类"surprisingly good" OR "actually good" OR "way better than I expected" ChatGPT OR Claude OR AI 2026← 质量评价类查询，可捕获学术 .edu 域名的研究文章（2026-08-01 新增，成功捕获 MIT Sloan 理财研究）
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.
- 12d ago First seen · 328 lines · 53 tokens per session scan C 28c62dc75a4e
ai-tips-tricks is a skill published in the GitHub repository 0xMulight/human-social-copy (23 stars, last pushed 1mo ago), licensed MIT. It adds 53 tokens to every session and 5,404 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
deai_humanize
A Chinese-language writing helper that measures signs of machine-like writing and rewrites text to sound more natural and personal.
humanize-automation
Human-like mouse, keyboard and scroll behavior for behavioral bot bypass.
mr-li-writer
A Chinese-language editing and writing workflow for turning ideas, source material, or rough topics into publishable content. It adapts the structure and wording to formats such as WeChat articles, Zhihu answers, Xiaohongshu posts, websites, and blogs.
21-day-self-interview
A 21-night guided self-reflection routine in which an agent asks three questions each night, remembers the answers, and reflects on them at key points.
api-doc-zh
An API documentation generator that turns an interface description or code into documentation in Chinese or English. The documentation can describe requests, responses, errors, permissions, and examples, using Markdown or OpenAPI 3.0 YAML.
competitive-analysis-zh
A Chinese-language tool for comparing a product with competitors. It creates a report covering the market, target users, business models, features, differences, weaknesses, and possible strategy.