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 LiHongwei-cn/lihongwei-cn --skill cheat-shootgit clone --depth 1 https://github.com/LiHongwei-cn/lihongwei-cnWrote 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/lihongwei-cn/lihongwei-cn/cheat-shoot)<a href="https://agentmods.dev/skills/lihongwei-cn/lihongwei-cn/cheat-shoot"><img src="https://agentmods.dev/badge/skills/lihongwei-cn/lihongwei-cn/cheat-shoot/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/lihongwei-cn/lihongwei-cn/cheat-shoot"><img src="https://agentmods.dev/badge/skills/lihongwei-cn/lihongwei-cn/cheat-shoot.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 | $0.00089 | $0.03337 |
| Opus 5 | $0.00044 | $0.01669 |
| Sonnet 5 | $0.00018 | $0.00667 |
| Haiku 4.5 | $0.00009 | $0.00334 |
Grade A, and why
cheat-shoot 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.
This is a copy
100% identical to cheat-shoot — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 214 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/cheat-shoot — 登记拍摄完成 + 建 video folder + (改稿则) 触发 v2 预测
把视频从"已写预测、未拍摄"状态推进到"已拍摄、未发布"状态。这一步:
- 建
videos/<同 id>/目录(之前没有的话) - 询问用户:"实际拍摄时用的稿子和
scripts/<id>.md一致吗?" - 算 diff——超过 V2_TRIGGER_THRESHOLD (默认 30%) → delegate 到
/cheat-predict — mode: v2在原 prediction 文件 append## 预测 v2段 - 把 video folder 加进 state.shoots 队列,buffer +1
cheat-shoot 自己不写预测内容——所有预测落盘逻辑在 cheat-predict。cheat-shoot 只负责检测改稿 + 派发。
为什么单独一个 skill:
- buffer 警戒系统需要明确区分"拍了" vs "发了"。视频可以批量拍(一天拍 5 条),分散发(每天发 1 条)
- "实际拍摄稿" ≠ "pre-shoot 草稿"是常态。这一步是把 diff 显式化、触发 v2 重判、采集"用户改稿 pattern"信号的入口
- v2 预测 vs v1 预测的差异本身就是 rubric 升级证据——比如 v1 给 ER=4,v2 给 ER=5(用户改稿改高了 hook 强度),就告诉 rubric "这个用户的 ER 阈值跟我现在公式不一致"
Overview
[用户:拍了 scripts/2026-05-04_abc123_停止期待.md]
↓
[Phase 0: 解析路径 + 验证 prediction 已存在]
↓
[Phase 1: 检查是否已登记(避免重复)]
↓
[Phase 2: 建 videos/<id>/ + 询问"实际拍摄稿一致吗?"]
↓
[Phase 3: 写 videos/<id>/script.md]
↓
[Phase 4: append state.shoots]
↓
[Phase 5: 输出 buffer 状态]
Constants
- REQUIRE_PREDICTION = true — 拍前必须先有 v1 prediction 文件
- V2_TRIGGER_THRESHOLD = 0.30 — normalize 后 char-level diff 超过 30% → 默认建议 v2 重判;低于 30% 询问用户是否仍要 v2
- DIFF_METRIC = char_levenshtein_normalized(默认)—— 通过
tools/diff_pct.py调用:先 normalize(去 markdown header / 分隔线 / 列表标记 / 装饰标点 / 折叠所有空白),再算 char-level Levenshtein / max(len_a, len_b)。preferred backendrapidfuzz,fallbackdifflib.SequenceMatcher(stdlib,永远可用)。旧版 line-level 在口语化转录场景误报严重(draft 长 markdown 句 vs whisper 转录的短断句,内容几乎不变但 line-level 算出 ~200% diff)—— PR #14 修复 - DIFF_METRIC=lines —— legacy fallback:当 python3 完全不可用或 tools/diff_pct.py 找不到时降级到
diff -u | grep '^[+-]' | wc -l算法
Inputs
| 必填 | 来源 |
|---|---|
<scripts-path-or-id> |
用户参数;缺失则询问 |
.cheat-state.json |
状态文件 |
scripts/*.md |
pre-shoot 草稿 |
predictions/*.md |
验证对应预测存在 |
Workflow
Phase 0:解析 + 验证
- 解析用户给的路径——支持几种形态:
- 完整路径
scripts/2026-05-04_abc123_停止期待.md - 简写
2026-05-04_abc123_停止期待 - id 简写
abc123→ globscripts/*_abc123_*.md找匹配
- 完整路径
- 验证
scripts/<id>.md存在:不存在 → 报错"找不到 pre-shoot 草稿" - 验证有对应 prediction
predictions/<同名>.md:- 不存在 → 拒绝登记,提示"先跑 /cheat-predict 写预测,否则违反盲预测原则——你不能拍完才写预测,那等于事后看了画面写"
- 存在 → 通过
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 · 214 lines · 89 tokens per session scan A 42fda83dc725
cheat-shoot is a skill published in the GitHub repository LiHongwei-cn/lihongwei-cn (5 stars, last pushed 1mo ago), licensed MIT. It adds 89 tokens to every session and 3,337 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to cheat-shoot, differing in 0 lines, and is treated as a copy.
Other skills, from other repositories
figure
Hybrid figure pipeline (Nano Banana raster + rembg background removal + TikZ vector assembly). Includes a TikZ template library covering quantum circuits (quantikz), Feynman diagrams (tikz-feynman), circuits (circuitikz), molecules (chemfig), 2D/3D plots (pgfplots), energy-level diagrams, phase-space trajectories…
illustration-style
Define an illustration style guide — visual language, colour usage, and application rules. Use when commissioning or standardising illustration. For icons, use icon-system (design-systems).
video-processing
Download, transcribe, analyze, and clip video content — vertical shorts, captions, thumbnails.
ai-provider-elevenlabs
ElevenLabs voice AI SDK patterns for TypeScript/Node.js -- text-to-speech, streaming, voice cloning, speech-to-speech, pronunciation control, and conversational AI.
operate-comfyui
Install, configure, operate, validate, and troubleshoot self-hosted ComfyUI through PawFlow relays, including host execution with local=true and service routing with relaylocal=true. Use for ComfyUI installation or repair, adding models or custom nodes, API workflow creation and submission, image/video/audio…
image-generation-specialist
Specialist guide for image generation and editing using available PawFlow media tools.