Lecture to Homework
vect-G/lecture-to-hw
Turns course slides, homework files, class code and earlier solutions into concise student-style Markdown answers, with optional subagents for solving and review.
“INTERNAL sub-agent for blind 7-dim rubric scoring. **NOT a user-facing skill — do NOT invoke from main conversation.** Called via Task tool by cheat-score / cheat-predict /…”
Install Blind Rubric Scorer Sub-Agent
$ npx skills add XBuilderLAB/cheat-on-content --skill cheat-score-blind -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install XBuilderLAB/cheat-on-content cheat-score-blind --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/XBuilderLAB/cheat-on-content.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/cheat-score-blind .claude/skills/cheat-score-blind && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "cheat-score-blind" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-score-blind into .claude/skills/cheat-score-blind/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-score-blind", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-score-blindType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add XBuilderLAB/cheat-on-content --skill cheat-score-blind -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install XBuilderLAB/cheat-on-content cheat-score-blind --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/XBuilderLAB/cheat-on-content.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/cheat-score-blind .agents/skills/cheat-score-blind && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "cheat-score-blind" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-score-blind into .agents/skills/cheat-score-blind/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-score-blind", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add XBuilderLAB/cheat-on-content --skill cheat-score-blind -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install XBuilderLAB/cheat-on-content cheat-score-blind --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/XBuilderLAB/cheat-on-content.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/cheat-score-blind .cursor/skills/cheat-score-blind && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "cheat-score-blind" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-score-blind into .cursor/skills/cheat-score-blind/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-score-blind", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/XBuilderLAB/cheat-on-content.git --path skills/cheat-score-blind--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add XBuilderLAB/cheat-on-content --skill cheat-score-blind -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install XBuilderLAB/cheat-on-content cheat-score-blind --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/XBuilderLAB/cheat-on-content.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/cheat-score-blind .gemini/skills/cheat-score-blind && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "cheat-score-blind" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-score-blind into .gemini/skills/cheat-score-blind/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-score-blind", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install XBuilderLAB/cheat-on-content cheat-score-blindInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add XBuilderLAB/cheat-on-content --skill cheat-score-blind -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/XBuilderLAB/cheat-on-content.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/cheat-score-blind .github/skills/cheat-score-blind && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "cheat-score-blind" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-score-blind into .github/skills/cheat-score-blind/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-score-blind", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add XBuilderLAB/cheat-on-content --skill cheat-score-blind -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install XBuilderLAB/cheat-on-content cheat-score-blind --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/XBuilderLAB/cheat-on-content.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/cheat-score-blind .opencode/skills/cheat-score-blind && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "cheat-score-blind" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-score-blind into .opencode/skills/cheat-score-blind/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-score-blind", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
cheat-score-blind- GitHub stars
- 7.2k
- Token cost
- ~2.1k tokens
- SKILL.md length
- 644 words
- Files
- 1
- Skills in repo
- 16
- Repo updated
- First seen
- Licence
- MIT
At a glance
- Works in 4 steps: ：边界自检 → ：按 rubric 打 N 维分 → ：返回严格 JSON → …
- SKILL.md covers Why this exists（绝不可省的背景）, 三 channel 模型, Inputs（唯一被允许的输入） and 禁止读取（hard list）, plus 5 more sections
- Calls python3
About this skill
“INTERNAL sub-agent for blind 7-dim rubric scoring. **NOT a user-facing skill — do NOT invoke from main conversation.** Called via Task tool by cheat-score / cheat-predict / cheat-bump to get a context-isolated score on a script. Receives ONLY script_path + rubric_notes_path; refuses any other input. Outputs strict JSON: 9 dimensions × {score 0-5, confidence enum, one-line reason}. **Hard refuses to Read** .cheat-state.json, predictions/*, retro 段, or anything that could leak post-publish data. This is channel B in the 3-channel calibration model (A=main, B=blind sub, C=cross-model).”
Blind Rubric Scorer Sub-Agent is a skill in XBuilderLAB/cheat-on-content (7.2k stars). Its SKILL.md is about 2.1k tokens. Licence: MIT.
Requirements
- Pre-approved tools (allowed-tools): Read, Glob, Grep
Workflow steps
4 steps, taken from the step headings in SKILL.md.
What it can do on your machine
Read from SKILL.md and the folder at commit 2d8211e. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves these tools, so the agent can use them without asking each time:
ReadGlobGrep
From allowed-tools in the SKILL.md frontmatter.
Runs code
Shell commands in SKILL.md call:
python3
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
Blind Rubric Scorer Sub-Agent loads about 2.1k tokens when it runs. Until then it costs about 152 tokens; SKILL.md has 644 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from XBuilderLAB/cheat-on-content at commit 2d8211e, republished under its MIT licence (© XBuilderLAB). 644 words, ~2,110 tokens.
Download SKILL.mdSave it as
.claude/skills/cheat-score-blind/SKILL.md (or your agent's skills folder).- name
- cheat-score-blind
- description
- INTERNAL sub-agent for blind 7-dim rubric scoring. **NOT a user-facing skill — do NOT invoke from main conversation.** Called via Task tool by cheat-score / cheat-predict / cheat-bump to get a context-isolated score on a script. Receives ONLY script_path + rubric_notes_path; refuses any other input. Outputs strict JSON: 9 dimensions × {score 0-5, confidence enum, one-line reason}. **Hard refuses to Read** .cheat-state.json, predictions/*, retro 段, or anything that could leak post-publish data. This is channel B in the 3-channel calibration model (A=main, B=blind sub, C=cross-model).
- allowed-tools
- Read, Glob, Grep
- argument-hint
- <script-path> <rubric-notes-path>
/cheat-score-blind — Channel B (blind scorer sub-agent)
⚠️ 这是子 agent，不是用户 skill。只能由
cheat-score/cheat-predict/cheat-bump通过 Task tool spawn。用户直接 trigger 没有意义——主对话已经被污染，调用 blind sub-agent 在主 context 里跑不构成隔离。
Why this exists（绝不可省的背景）
cheat-on-content 的 7/9 维打分原本 inline 在主对话——但主 Claude 已经看过：
- 用户对话历史（含偶然提到的播放数 / 评论 / 情绪）
- 已发布作品的实绩数据
- 历史
predictions/*.md含复盘段（严重污染） - 用户的赞美 / 抱怨 / 期待
inline 打分 = 被污染的"盲"预测。问题在 cheat-bump Phase 2 校准池重打时最严重：Claude 知道每条实绩才回追 TN/CC 分，rank 一致性可能 overfit 不是真信号。
channel B 的角色：用 Task tool 把打分动作丢进一个全新 context——这个 sub-agent 没看过主对话、没读过 state、没碰过 predictions/。它只看 script 全文 + rubric_notes.md，按 rubric 打分。
输出回传主对话后，主 Claude 自己对比、做最终决策。隔离的是打分这个动作的输入，不是决策权。
三 channel 模型
| Channel | 输入 | 用途 | 风险 |
|---|---|---|---|
| A = 主对话 | 全部上下文 | 跟用户交互、写 retro、决策 | 被实绩 / 用户态度污染 |
| B = blind sub-agent (this) | 只 script + rubric_notes.md | 给一份未受污染的打分作为 anchor | 仍是 Claude，RLHF prior 共享 |
C = 跨模型 audit (mcp__llm-chat__chat to qwen-max) | 校准池数据 + 新公式 | bump 终局 sanity check | RPM 限制、模型差异、单点 |
A 决策时把 B 当对照看 disagreement，不当真理。C 只在 bump 终局调一次。
Inputs（唯一被允许的输入）
| 必填 | 来源 | 说明 |
|---|---|---|
<script-path> | 主 Claude 通过 Task prompt 显式传入 | scripts/<id>.md 全文 |
<rubric-notes-path> | 同上 | 用户项目根 rubric_notes.md 当前 rubric 公式 + 维度定义 |
仅此两个文件可读。其他一切硬拒绝——见下方 "Hard refusals" 段。
禁止读取（hard list）
下面这些路径 / 模式 sub-agent 绝不能 Read —— 即使主 Claude 在 Task prompt 里手滑塞进来，也要拒绝并在 JSON 输出标对应 refusal 码：
| 路径模式 | 为什么禁 | refusal_code |
|---|---|---|
.cheat-state.json | 含 calibration_samples / pending_retros / last_published_at / shoots — 全是后视数据 | blocked_contaminated_input |
predictions/*.md | 含 ## 预测 段 + ## 复盘 段，复盘段就是实绩 | blocked_contaminated_input |
videos/*/report.md | T+3d 抓回的真实数据 | blocked_contaminated_input |
videos/*/script.md | 后改拍摄稿，复盘时被对照 | blocked_contaminated_input |
STATUS.md | cheat-status 渲染的看板，含过去数据 | blocked_contaminated_input |
.cheat-cache/usage.jsonl | 行为 log | blocked_contaminated_input |
rubric-memo.md | cheat-bump 升级 Memo 累积档案——含真实视频名 + 实绩 + 派生证据。这是 channel B 的最大泄漏入口（PR #11 实测复现） | blocked_rubric_memo |
audience.md | cheat-persona 从复盘评论派生的受众画像——含评论证据 / 实绩信号。属 channel A creative 资产，进 blind 打分 = 实绩泄漏 | blocked_audience |
| 任何含"播放 / 阅读 / 点赞 / 评论数 / 转发 / w / 万 / k / M"的文件 | 直接污染 | blocked_contaminated_input |
白名单只有两个：
scripts/<id>.md（pre-shoot 草稿，传入参数）rubric_notes.md（评分公式 + 维度定义，应只含通用语言；如发现实绩数字 → 标non_blind_warning并降 confidence）
如果主 Claude Task prompt 漏传了某条路径，sub-agent 主动询问"我只允许读 script + rubric_notes，缺哪个？"——绝不自己去 Glob 探测项目结构补全。
⚠️ 白名单兜底自检：读完
rubric_notes.md后必跑grep -E '\\d+\\s*[wWmMkK万]|播放|实绩|实际'——命中 → 标self_check.any_contamination_signal: true+refusal: "non_blind_warning"，所有维度 confidence 降 medium 并把违禁 snippet 摘抄进 contamination_note 字段。仍输出 dimensions 让主 Claude 知道发生了什么——拒绝输出比误判更糟，但要诚实标注。
Workflow
Phase 0：边界自检
- 解析 Task prompt 拿
<script-path>和<rubric-notes-path> - 校验路径符合白名单——不在
scripts/下的 .md → 拒绝（除非主 Claude 显式说明"这是临时草稿临时路径，标non_standard_path: true"） - Read
<rubric-notes-path>→ 解析当前 rubric_version + 维度数量（7 或 9）+ 公式 - Read
<script-path>→ 拿到 script 全文 + 字数
⚠️ 不要做的事：
- 不要为了"看看用户做啥账号"去 Read
benchmark.md—— benchmark 是 Channel A 的 context，不属于本 sub-agent - 不要为了"看看历史风格"去 Glob
predictions/—— 那是污染源 - 不要去 Read
.cheat-state.json看 calibration 进度 —— 你完全不需要知道主 Claude 跑了多少篇
Phase 1：按 rubric 打 N 维分
按 rubric_notes.md 当前 rubric 公式：
- v0：7 维等权（ER / SR / HP / QL / NA / AB / SAT）—— 默认起步
- v1：用户校准过的（权重不同）
- v2 / v2.1 / ...：含 MS / TS 等新增维度（9 维）
对每个维度：
- 给一个 0-5 整数分
- 给一个 per-dim confidence enum：
high | medium | low- high：稿子里有直接证据（一句话指向该维度）
- medium：可推断但需要解释
- low：稿子信号太弱，纯估
- 给一行 理由 ≤ 30 字，必须引用稿子里具体词或场景
不算 composite——composite 是公式行为，主 Claude 用回传的维度分自己算。
Show full SKILL.md (252 more words)Show less
Phase 2：返回严格 JSON
输出只能是一个有效 JSON。所有 markdown 解释都封禁——主 Claude 要的是结构化数据回主 context 解析。
{
"subagent_version": "v1",
"rubric_version": "v2",
"script_path": "scripts/2026-05-04_abc123_短title.md",
"script_hash": "<sha256:12 of script content>",
"scored_at": "<ISO 8601 +08:00>",
"dimensions": {
"ER": { "score": 4, "confidence": "high", "reason": "PPT加油猫猫开头—具象画面，情绪反差强" },
"SR": { "score": 3, "confidence": "medium", "reason": "AI焦虑是议题但非热点对峙" },
"HP": { "score": 5, "confidence": "high", "reason": "首句\"第七页大屏中央 加油猫猫\"具象反差" },
"QL": { "score": 5, "confidence": "high", "reason": "\"加油猫猫救了我一命\"双关金句" },
"NA": { "score": 4, "confidence": "medium", "reason": "单线反思+收束，清晰但不复杂" },
"AB": { "score": 4, "confidence": "medium", "reason": "一人公司题但AI焦虑普适" },
"SAT": { "score": 2, "confidence": "high", "reason": "共情调，几乎无讽刺" }
},
"input_status": {
"rubric_notes_read": true,
"script_read": true,
"any_other_file_read": false
},
"self_check": {
"saw_play_numbers": false,
"saw_comments": false,
"saw_retro_segment": false,
"any_contamination_signal": false
},
"refusal": null
}refusal != null 的合法值：
"blocked_contaminated_input"：Task prompt 传了禁读路径（state / predictions / videos / 等）"blocked_rubric_memo"：Task prompt 传了rubric-memo.md（bump 升级档案，含实绩）"blocked_audience"：Task prompt 传了audience.md（受众画像，含评论派生的实绩信号）"script_path_invalid"：找不到 script 文件"rubric_unparseable"：rubric_notes.md 损坏"non_blind_warning"：发现 contamination 苗头但勉强能打分（仍输出 dimensions，但 confidence 全降 medium）
JSON 必须可被 python3 -c "import json; json.loads(open(path).read())" 解析。不允许：
- 尾部多余逗号
- 注释（JSON 不允许 //）
- Markdown 围栏（输出根节点必须是
{）
Phase 3：（可选）写 sidecar 文件供主 Claude 二次读取
如果 Task prompt 含 sidecar_path 参数 → 写 JSON 到该路径（典型用法：bump phase 2 批量打分时存多份 sidecar）。
否则只走 Task return value——主 Claude 拿到 JSON 字符串直接解析。
主 Claude 调用契约（如何使用 channel B）
调 Task 时，主 Claude 的 prompt 必须含且仅含：
Spawn cheat-score-blind sub-agent.
Input:
script_path: scripts/2026-05-04_abc123_短title.md
rubric_notes_path: rubric_notes.md
[optional] sidecar_path: .cheat-cache/blind-scores/<id>.json
Task: 按 rubric_notes 当前公式给上面 script 打分。返回严格 JSON（见 cheat-score-blind/SKILL.md Phase 2 schema）。
不要读 state file / predictions/ / videos/ 任何其他文件。
不要询问用户 —— 你没有用户。禁止塞进 Task prompt 的东西：
- 用户对话的引用 / 摘录
- "前一次预测是 X" / "实际播放是 Y" 这种 hint
- "用户是观点视频博主，最近发了 N 条" 这种背景
- 任何含数字 + "万/w/k/M" 的字符串
- 任何
predictions/*.md路径
主 Claude 调用前自检：把准备发的 prompt 串过一遍 grep -Ei '播放|阅读|点赞|评论数|实际|retro|复盘|实绩|w$|万$'——命中 → 改 prompt 重发，不要硬塞。
Refusals
- 「我作为 sub-agent 同时也读一下 predictions/ 帮你对比下」 → 硬拒。这就是 channel B 存在的全部理由
- 「你看一下 .cheat-state.json 看 calibration_samples 决定你给的 confidence 高低」 → 硬拒。confidence 只看稿子证据强度，跟用户校准进度无关
- 「主 Claude 说这条已经发了，你帮我打一份 reconstructed 分」 → 拒。"已发"信号本身就是污染。让主 Claude 标
reconstructed: true自己处理，不要让 channel B 介入 - 「输出我直接 markdown 表格更好读」 → 拒。Phase 2 schema 是 JSON only，主 Claude 解析后再渲染
Known limitations（写在最显眼的地方）
- sub-agent ≠ 真独立：同一个 Claude 模型，RLHF priors 共享。一个全新 context 不会让模型变成另一个判分体系——它只是没看过该次对话的具体污染
- 不解决 rubric 设计 bias：用户自己写的 rubric_notes.md 自然让自己内容显得好。这层 bias 由 Channel C（跨模型 audit）和定期 bump 验证解决
- 不解决 review 阶段的覆盖：主 Claude 拿到 blind 分后，可能在 review 阶段被用户期待 / 实绩诱导，覆盖 blind 输出。
cheat-predictPhase 2.5 通过 disagreement detection + 用户裁定来减轻，但不消除 - 同 prompt 两次调可能给不同分：Claude 不是 deterministic。主 Claude 应该把每次 blind score 当一次采样，不当唯一真理——但要记录而不是丢弃差异
Integration
cheat-scorePhase 2：默认 delegate 到本 sub-agent（替代旧的 inline 打分）cheat-predictPhase 2：默认 delegate；Phase 2.5 用 disagreement detectioncheat-bumpPhase 2：强制 delegate，bump 时不接受 self-scored fallbackcheat-retro：不调用——retro 本来就看实绩，blind 无意义
© XBuilderLAB, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
Just SKILL.md in skills/cheat-score-blind of XBuilderLAB/cheat-on-content.
Open the folder on GitHubat commit 2d8211e
Compare with similar skills
Blind Rubric Scorer Sub-Agent next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Blind Rubric Scorer Sub-Agent this skillXBuilderLAB/cheat-on-content | 7.2k | — | ~2.1k | Automated safety check: Pass | MIT | |
| Lecture to Homeworkvect-G/lecture-to-hw | 127 | — | ~729 | Automated safety check: Pass | MIT | |
| Woo AI Smokewoocommerce/woocommerce-ios | 358 | 1 repos | ~7.4k | Automated safety check: Notes | GPL-2.0 | |
| Canvas Awkward SyntaxX-isdoingreat/canvas-pilot | 125 | — | ~15k | Automated safety check: Notes | AGPL-3.0 | |
| Canvas GenericX-isdoingreat/canvas-pilot | 125 | — | ~7.5k | Automated safety check: Notes | AGPL-3.0 | |
| ArenaJakeschincariol/arena-skill | 312 | — | ~4.8k | Automated safety check: Pass | MIT |
Similar skills
Woo AI Smoke
woocommerce/woocommerce-ios
Evaluate WooAIAssistant against a structured scenario suite with hard invariants + LLM-as-judge rubric scoring.
Canvas Awkward Syntax
X-isdoingreat/canvas-pilot
Direct-commit syntactically-marked rewrite humanizer. An agent skill from X-isdoingreat/canvas-pilot.
Canvas Generic
X-isdoingreat/canvas-pilot
Fallback runtime-designed handler for Canvas assignments that don't fit any of the 5 specific skills (canvas-ics33 / canvas-reading-annotation / canvas-essay / canvas-zybooks / canvas-inside).
Arena
Jakeschincariol/arena-skill
Make 100 versions of Claude fight to the death over one task.
Claude Code Advanced Guide
2025Emma/vibe-coding-cn
A Chinese-language tutorial on advanced Claude Code use: its core tools, REPL, slash commands, workflows, permissions, memory, MCP, hooks and sub-agents.
More from XBuilderLAB/cheat-on-content
All 16 skills in this repoRubric Bump Proposer
XBuilderLAB/cheat-on-content
Proposes and applies upgrades to a content-scoring rubric: a full formula bump with blind re-scoring and a cross-model audit, or a lighter bucket-boundary recalibration.
Cheat-on-Content Setup
XBuilderLAB/cheat-on-content
Onboards a new user to the cheat-on-content workflow with a short question flow, creating the project scaffolding and optionally importing past video history.
Benchmark Account Importer
XBuilderLAB/cheat-on-content
Imports scripts and engagement numbers from an account you want to emulate, then extracts content patterns and starting scoring signals from them.
cheat-on-content State Migrator
XBuilderLAB/cheat-on-content
Upgrades an older .cheat-state.json to the current schema version by applying migration files in order, with dry-run, backup and stop-on-failure behavior.
Cheat on Content Calibration
XBuilderLAB/cheat-on-content
Turns content creation into a calibrated loop of scoring, blind prediction, post-publish review and rubric evolution, with a built-in rubric for opinion videos.
Audience Persona Builder
XBuilderLAB/cheat-on-content
Builds or refreshes an account's audience profile from the comments in its post retrospectives and writes it to audience.md for later topic and script work.
Related
Categories
Questions about Blind Rubric Scorer Sub-Agent
How do I install Blind Rubric Scorer Sub-Agent in Claude Code?
Run `npx skills add XBuilderLAB/cheat-on-content --skill cheat-score-blind -a claude-code`. Or copy the skill folder (skills/cheat-score-blind in XBuilderLAB/cheat-on-content) into .claude/skills/cheat-score-blind in your project. Claude Code loads it when a task matches its description.
How do I install Blind Rubric Scorer Sub-Agent in Codex?
Run `npx skills add XBuilderLAB/cheat-on-content --skill cheat-score-blind -a codex`. Or copy the skill folder (skills/cheat-score-blind in XBuilderLAB/cheat-on-content) into .agents/skills/cheat-score-blind in your project. Codex loads it when a task matches its description.
Can I use Blind Rubric Scorer Sub-Agent in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add XBuilderLAB/cheat-on-content --skill cheat-score-blind -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/cheat-score-blind, .gemini/skills/cheat-score-blind, .github/skills/cheat-score-blind and .opencode/skills/cheat-score-blind in your project.
What does Blind Rubric Scorer Sub-Agent need to run?
Going by SKILL.md and its folder, Blind Rubric Scorer Sub-Agent needs the command-line tools its instructions call (python3). Its frontmatter pre-approves these tools: Read, Glob, Grep.
Does Blind Rubric Scorer Sub-Agent access the network?
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Is Blind Rubric Scorer Sub-Agent safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does Blind Rubric Scorer Sub-Agent use?
Blind Rubric Scorer Sub-Agent is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Blind Rubric Scorer Sub-Agent use?
About 2.1k tokens (SKILL.md is roughly 8.4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
What are the alternatives to Blind Rubric Scorer Sub-Agent?
Skills that share tags, products or a category with Blind Rubric Scorer Sub-Agent: Lecture to Homework (vect-G/lecture-to-hw, 127 stars), Woo AI Smoke (woocommerce/woocommerce-ios, 358 stars), Canvas Awkward Syntax (X-isdoingreat/canvas-pilot, 125 stars) and Canvas Generic (X-isdoingreat/canvas-pilot, 125 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Blind Rubric Scorer Sub-Agent?
XBuilderLAB (a GitHub organization) maintains it in XBuilderLAB/cheat-on-content, which has 7,225 GitHub stars. The repository holds 16 skills in this directory. The repository was last updated on October 5, 2026.
Source: XBuilderLAB/cheat-on-content on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.