Schedule Social
indranilbanerjee/digital-marketing-pro
Schedule social posts to Twitter/X, Instagram, LinkedIn, TikTok, YouTube, and Pinterest through connected platform MCPs — generating per-platform copy variations, hashtag mixes, media-spec…
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.
SKILL.md written in Chinese; this summary is our English description.
$ npx skills add XBuilderLAB/cheat-on-content --skill cheat-bump -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install XBuilderLAB/cheat-on-content cheat-bump --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-bump .claude/skills/cheat-bump && 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-bump" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-bump into .claude/skills/cheat-bump/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-bump", 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-bumpType 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-bump -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install XBuilderLAB/cheat-on-content cheat-bump --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-bump .agents/skills/cheat-bump && 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-bump" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-bump into .agents/skills/cheat-bump/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-bump", 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-bump -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install XBuilderLAB/cheat-on-content cheat-bump --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-bump .cursor/skills/cheat-bump && 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-bump" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-bump into .cursor/skills/cheat-bump/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-bump", 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-bump--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-bump -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install XBuilderLAB/cheat-on-content cheat-bump --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-bump .gemini/skills/cheat-bump && 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-bump" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-bump into .gemini/skills/cheat-bump/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-bump", 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-bumpInstalls 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-bump -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-bump .github/skills/cheat-bump && 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-bump" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-bump into .github/skills/cheat-bump/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-bump", 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-bump -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-bump --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-bump .opencode/skills/cheat-bump && 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-bump" agent skill from https://github.com/XBuilderLAB/cheat-on-content/tree/main/skills/cheat-bump into .opencode/skills/cheat-bump/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cheat-bump", 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.
cheat-bumpProposes 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.
Two modes exist and are never mixed in one run. A full rubric bump (--propose) changes the formula, its dimensions or weights, and must follow a five-step validation protocol with a cross-model review. A bucket-only run (--bucket-only) re-derives bucket boundaries from data without changing the rubric formula or requiring review, for example when an account has grown and the buckets no longer fit.
A full bump first checks readiness, expands the proposal into a complete equation, and requires the calibration pool to be re-scored by the cheat-score-blind sub-agent; self-scored fallbacks are not accepted. The new ranking must agree with actual results at a fixed threshold of 0.8 (four of five), an external model audits the change, and nothing is applied until you confirm with yes, bump. Inputs are your rubric_notes.md, the predictions folder and a .cheat-state.json state file.
9 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit 2d8211e. It shows what the files ask for, not the result of running them.
Pre-approves these tools, so the agent can use them without asking each time:
Bash(*)ReadWriteEditGlobGrepSkillTaskmcp__llm-chat__chatFrom allowed-tools in the SKILL.md frontmatter.
No scripts in the folder and no shell commands in SKILL.md (its code samples are markdown and json).
From the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Rubric Bump Proposer loads about 3.5k tokens when it runs. Until then it costs about 73 tokens; SKILL.md has 992 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.
The automated check noted patterns worth knowing about, such as sudo or a known installer.
allowed-tools: Bash(*), Read, Write, Edit, Glob, Grep, Skill, Task, mcp__llm-chat__chatAutomated 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.
The full file from XBuilderLAB/cheat-on-content at commit 2d8211e, republished under its MIT licence (© XBuilderLAB). 992 words, ~3,472 tokens.
.claude/skills/cheat-bump/SKILL.md (or your agent's skills folder).两种模式:
| 模式 | 触发 | 做什么 | 验证强度 |
|---|---|---|---|
| 完整 rubric bump | --propose "<新公式>" | 改公式 / 维度 / 权重 | 5 步 + 跨模型审核(强制) |
| bucket-only 重校 | --bucket-only | 只重新派生 bucket 边界 | 数据自动派生,无审核 |
完整 rubric bump 严格遵守 shared-references/bump-validation-protocol.md 的 5 步。bucket-only 走轻量路径——见下方 Phase B。
入口:用户触发 /cheat-bump
↓
[Phase A0: 检测调用模式]
↓
├─ --bucket-only → [Phase B: 轻量 bucket 重校]
└─ --propose → [Phase 0~6: 完整 rubric bump]读用户参数:
--bucket-only → 走 Phase B(轻量重校)--propose "<...>" → 走 Phase 0~8(完整 rubric bump)如果用户说"我觉得 ER 太低了想调"→ 是 --propose 路径。
如果用户说"我账号长大了,bucket 不准了"→ 是 --bucket-only 路径。
两条路径不能混调——一次操作只做一种事。
[用户:升级 rubric --propose "ER×1.5→2.0,砍 NA,加 MS"]
↓
[Phase 0: 前置门槛检查]
↓
[Phase 1: 写出新公式完整方程]
↓
[Phase 2: 校准池全量重打分]
↓
[Phase 3: 计算排序一致性]
↓
[Phase 4: 跨模型独立审核(强制)]
↓
[Phase 5: 落地 + cleanup pass]
↓
[Phase 6: 更新所有校准样本的 prediction 文件底部追加 Re-scored 行]| 必填 | 来源 |
|---|---|
--propose 文本 | 用户参数;缺失则询问 |
rubric_notes.md | 用户项目根 |
predictions/*.md 全量 | 校准池数据 |
.cheat-state.json | 状态 |
按 bump-validation-protocol.md 的"何时禁止"段,逐项检查:
| 检查 | 失败处理 |
|---|---|
| 校准池总样本数 vs 观察强度 | Claude 判断——按 READINESS_HEURISTIC:默认 ≥5 样本但允许特例(强反例 / 强模因)。如不满足默认,Claude 必须显式说明为什么仍然提议 bump("虽然只 N=3 样本,但 X 这条出现 composite Y vs 实绩 Z,这是 W 倍偏差"),让用户审视 |
| 上次 bump 距今的新校准数 vs 观察成熟度 | Claude 判断——默认建议 ≥3 篇新样本,但如果连续 3 篇都强证据指向同一方向 → 不必再等 |
in_progress_session == null | 拒绝:"你有 in-progress 预测未完成。先走完那条流程或清掉 state" |
| 触发条件成立(系统性偏差 / 跨样本新观察 / 新维度证据足) | 警告但不阻塞——询问用户为什么现在 bump |
通过 → 进入 Phase 1。
不能只接受用户的简短描述。把它展开为完整方程:
当前:v2 composite = (ER×1.5 + SR×1.5 + HP×1.5 + QL + NA + AB + SAT) / 8.5 × 2.0
提议:v2.1 composite = (ER×2.0 + HP×1.5 + MS×1.5 + QL + SR + TS + SAT) / 9.0 × 2.0
变化总结:
- ER ×1.5 → ×2.0(升)
- SR ×1.5 → ×1.0(降)
- 新增 MS ×1.5(Memetic Shareability)
- 新增 TS ×1.0(Topic Shareability)
- 删除 NA(与 HP 重叠)
- 删除 AB(被 TS 替代)
- 归一化常数 8.5 → 9.0
- 公式总维度数:7 → 7(净变化 0)如果用户的提议含糊(如"ER 权重提一点")→ 询问具体数值,禁止自己猜。
Glob predictions/*.md 中所有有完整复盘段的文件 → 校准池。
bump 是工具最高风险动作——所有重打必须走 cheat-score-blind sub-agent。inline 重打 = 主 Claude 已经看过实绩,rank 一致性变成 overfit 而非真信号。
/cheat-predict 有 --skip-blind flag,但 /cheat-bump 没有。如果 Task tool 不可用 → abort bump,向用户报告"先解决 Task tool 再 bump"scripts/<id>.md 路径(从 Script Path header 字段)Script Hash 一致;不一致 → 警告(script 改过了)但仍 spawn sub-agentSpawn cheat-score-blind sub-agent.
Input:
script_path: <prediction header 的 Script Path>
rubric_notes_path: rubric_notes.md
sidecar_path: .cheat-cache/bump-rescores/<prediction-id>.json
Task: 按 rubric_notes 当前公式(已是新版 vN+1)给 script 打分。
返回严格 JSON。写 sidecar 文件用于 bump 主流程批量读取。
不要读 state file / predictions/ / videos/ 任何其他文件。
不要询问用户 —— 你没有用户。
不要读这份 prediction 文件本身 —— 你只看 script + rubric。.cheat-cache/bump-rescores.json(汇总)。每条 entry 标 blind: true —— bump phase 5 cleanup 时把这个字段连同新分数写到 prediction 文件的 Re-scored under v<N+1> 行即使走 sub-agent,仍有两类残余 contamination 要在 bump report 里诚实标注:
| 类型 | 来源 | 标注字段 |
|---|---|---|
| 模型 prior contamination | sub-agent 仍是 Claude,RLHF 共享 | model_prior_warning: true(默认 true,不可关) |
| 用户自己 rubric design bias | rubric_notes.md 是用户写的,自然 fit 自己内容 | rubric_self_designed: true(默认 true,不可关) |
这两条提示用户 channel C(跨模型 audit)的不可省。bump 报告末尾必印:"上面的 rank 一致性是 channel A 内的一致性。最终决策必须等 channel C audit 通过。"
| 症状 | 处理 |
|---|---|
| 某条 prediction 的 script 文件不见了 | sub-agent skip 该条,主流程汇总报告"N 条因 script 缺失被排除"。如剩余有效池 < MIN_SAMPLES → abort bump |
sub-agent 返回 refusal != null | 重发 Task 最多 3 次;仍败 → 该条标 rescore_failed: true 排除出校准池 |
| Task tool 整个不可用 | abort bump,提示用户"Task tool 是 bump 的硬依赖。如真的离线环境,跑 /cheat-bump --bucket-only 走轻量分支" |
| sub-agent 输出含 contamination_signal | 标 suspicious: true 但不排除——bump report 末尾列这些可疑条目让用户审 |
每个样本:
new_composite_rank: 用新公式排序的 rank
actual_plays_rank: 用实际播放排序的 rank
delta: |new_rank - actual_rank|
输出对照表:
| 样本 | composite (v2) | composite (v2.1) | rank (new) | actual | rank (actual) | delta |
|---|---|---|---|---|---|---|
| 仓鼠 | 9.41 | 9.55 | 1 | 124.8w | 1 | 0 |
| 停止期待 | 8.24 | 9.11 | 2 | 71.1w | 2 | 0 |
| 老板废话 | 7.65 | 8.11 | 4 | 39.6w | 3 | 1 |
| 求职悖论 | 8.47 | 7.56 | 5 | 16.8w | 4 | 1 |
| 谁问你了 | 8.24 | 7.00 | 6 | 11.7w | 5 | 1 |
排序一致性:4/5 在 |delta| ≤ 1
Pairwise no-regression:旧公式做对的所有 pair 在新公式下未颠倒 ✓判定:
THRESHOLD 写死在协议里——不允许临时调低(那本身是另一个需要 bump 的元决策)。
CROSS_MODEL_AUDIT=true(默认):
调用 mcp__llm-chat__chat:
prompt:
你是一个独立审稿人。下面是一个内容创作者准备升级的 rubric 公式。
请独立判定两件事:
1. 排序一致性:新公式给样本的排序与实际表现排序,是否真的在 ≥80% 样本上一致?
2. 解释力:新公式相比旧公式,是否更好地解释了校准池的实绩分布?
数据:
旧公式:(ER×1.5 + SR×1.5 + HP×1.5 + QL + NA + AB + SAT) / 8.5 × 2.0
新公式:(ER×2.0 + HP×1.5 + MS×1.5 + QL + SR + TS + SAT) / 9.0 × 2.0
校准池:
[Phase 2 重打表的完整 JSON]
排序对照:
[Phase 3 表格的完整 JSON]
输出格式:
- 判定:PASS 或 REJECT
- 理由:≥100 字
- 关键风险:[如有,列出新公式的潜在问题]收到外部 LLM 回复 → 解析判定。
判定逻辑:
CROSS_MODEL_AUDIT=false,state file 标 last_bump_self_audited: trueCROSS_MODEL_AUDIT=false:
通过审核后,REQUIRE_CONFIRM=true → 询问用户:"新公式 PASS 本地与外部审核。最后确认:执行 bump 落地?这会修改 rubric_notes.md + rubric-memo.md 并删除若干已被吸收的观察。回答 'yes, bump' 才执行。"
用户确认后:
rubric_notes.md(只放通用语言,不含视频名 / 实绩)**当前版本**: vN+1**Last bumped at**: <ISO 8601>**Upgrade memos**: 见 [rubric-memo.md](rubric-memo.md)(指针,不复制 Memo 内容)rubric-memo.md(append 模式,不覆盖历史)按 bump-validation-protocol.md Step 5 + templates/rubric-memo.template.md 格式 append 一段 Memo 到文件末尾:
绝不覆盖 rubric-memo.md 已有内容——bump memo 按时间顺序累积。
在 rubric_notes.md 内执行(不动 rubric-memo.md):
rubric_notes.md 全文,确保读者能在 60 秒内理解当下规则——超出 600 行触发额外清算rubric_notes.md 跑 grep -E '\\d+\\s*[wWmMkK万]|播放|实绩|实际' → 如有命中 → abort bump + 回滚,提示用户"rubric_notes.md 写入了违禁内容(实绩 / 播放数)"。这些内容应在 rubric-memo.md,不在 rubric_notes.md对每个校准样本的 prediction 文件,底部追加(不动预测段、不动复盘段):
---
**Re-scored under v2.1 on 2026-05-04**: composite=8.24 → 9.11 (blind: true)
(rubric bump 时全量重算,由 cheat-score-blind sub-agent 独立打分;详见 rubric-memo.md 的 v2 → v2.1 升级 Memo)blind: true 字段必填——告诉未来读这条记录的人"这是 channel B 隔离打分,不是主 Claude 自评"。如果某条 prediction 在 Phase 2 因 sub-agent 失败被排除 → 不会有 Re-scored 行(保持原样)。
用 Edit 工具,匹配每个文件的最末尾。
{
"rubric_version": "v2.1",
"last_bump_at": "<ISO timestamp>",
"last_bump_self_audited": false,
"consecutive_directional_errors": [],
"calibration_samples_at_last_bump": <current value>
}清空 consecutive_directional_errors——新 rubric 重新计数。
✅ Rubric 已升级 v2 → v2.1
变化:
- ER ×1.5 → ×2.0
- SR ×1.5 → ×1.0
- 新增 MS / TS
- 删除 NA / AB
校准池重打:5/5 通过排序检查(4/5 一致 + 0 pairwise 回归)
跨模型审核:✅ PASS
Cleanup pass:删除观察 D 和 E(已吸收为 QL 重定义和 MS 维度)
下一篇预测起按 v2.1 公式打分。
所有历史预测文件已追加 Re-scored 标记。/cheat-bump --bucket-only [--scheme ratio|absolute|percentile]
与完整 bump 的本质区别:bucket 边界不是规则的一部分,是数据派生量。重新派生它不需要跨模型审核——派生算法是确定性的,没有"判断"成分。
| 算法 | 适用 | 边界派生方式 |
|---|---|---|
ratio(默认 N=1-4) | 小样本 | 上一篇 / 最近 3 篇中位数 × {0.3 / 1 / 3 / 10 / 30} |
absolute(默认 N=5-9) | 中等样本 | 校准池中位数 × {0.3 / 1 / 3 / 10 / 30},固定边界 |
percentile(默认 N≥10) | 大样本 | 校准池实绩 percentile {30 / 60 / 85 / 95 / 100} |
--scheme 参数允许用户显式覆盖默认:
--scheme ratio 强制用 ratio(即使 N≥5)--scheme absolute 强制用 absolute--scheme percentile 强制用 percentile(要求 N≥3,否则报错)未指定 --scheme → 按上表自动派生。
旧设计有
bucket_schemestate 字段——v1.1 删了。所有 skill 实时按 calibration_samples 派生算法,不需要持久化"当前用哪个"。这避免了"切换 scheme 后忘了同步"的状态不一致问题。
读 predictions/*.md 中所有有 actual_plays 的样本。
ratio 模式:
baseline = median(最近 3 篇 actual_plays)
buckets = {
"退步": (-inf, baseline * 0.3),
"持平": (baseline * 0.3, baseline * 1),
"命中": (baseline * 1, baseline * 3),
"小爆": (baseline * 3, baseline * 10),
"大爆": (baseline * 10, +inf),
}absolute 模式:
baseline = median(全部校准池 actual_plays)
buckets = {
"底部": (-inf, baseline * 0.3),
"基础盘": (baseline * 0.3, baseline * 1),
"命中": (baseline * 1, baseline * 3),
"爆款": (baseline * 3, baseline * 10),
"现象级": (baseline * 10, +inf),
}percentile 模式:
sorted_plays = sorted(全部校准池 actual_plays)
buckets = {
"底部": ≤ p30,
"基础盘": p30 - p60,
"命中": p60 - p85,
"小爆": p85 - p95,
"大爆": ≥ p95,
}当前 bucket scheme: ratio
proposed scheme: absolute
baseline: 4.2w 中位数(基于 5 篇校准样本)
新边界:
- 底部: < 1.3w
- 基础盘: 1.3w - 4.2w
- 命中: 4.2w - 12.6w
- 爆款: 12.6w - 42w
- 现象级: > 42w
派生说明:
- 5 篇实绩:1.5w / 3.8w / 4.2w / 5.6w / 18w
- 中位数 4.2w,新桶按 ×{0.3, 1, 3, 10} 派生
确认应用?(yes / no)用户确认后:
rubric_notes.md 的 "Bucket 方案" 段,替换为新表.cheat-state.json 的 baseline_plays 字段(bucket scheme 不持久化——下次 cheat-predict 实时派生)rubric_notes.md 的 bucket 段顶部追加一行变更记录:v2 buckets recalibrated on YYYY-MM-DD: scheme=absolute, baseline=4.2w (基于 N=10 个样本)下一次 /cheat-predict 起按新 bucket 派生。历史 prediction 文件里的 bucket 标签不重算——bucket 是预测时的语义判断,事后改写会破坏盲度。
CROSS_MODEL_AUDIT=false 显式设置/cheat-retro 检测到 ≥3 同向偏差 → 提议跑 /cheat-bumpmcp__llm-chat__chat(如配置)+ Task tool(spawn cheat-score-blind)rubric_notes.md(结构性更新,绝不写真实视频名 / 实绩)rubric-memo.md(新——append Memo 全文,含证据 + 派生证据)predictions/*.md(追加 Re-scored 行,不动预测段).cheat-state.json/cheat-predict 自动按新 rubric_version 打分© XBuilderLAB, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Just SKILL.md in skills/cheat-bump of XBuilderLAB/cheat-on-content.
Open the folder on GitHubat commit 2d8211e
Rubric Bump Proposer 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 |
|---|---|---|---|---|---|---|
| Rubric Bump Proposer this skillXBuilderLAB/cheat-on-content | 7.2k | — | ~3.5k | Automated safety check: Notes | MIT | |
| Schedule Socialindranilbanerjee/digital-marketing-pro | 855 | 1 repos | ~3.4k | Automated safety check: Pass | MIT | |
| 11 Channel Setup Globalminhnv0807/ai-business-skills | 608 | — | ~3.2k | Automated safety check: Pass | MIT | |
| Running Marketing Campaignsnicepkg/ai-workflow | 285 | — | ~2.7k | Automated safety check: Pass | MIT | |
| SEO AI Search Share Of Voiceseranking/seo-skills | 160 | — | ~1.1k | Automated safety check: Pass | MIT | |
| Competitor Analysissocial-media-skills/skills | 125 | — | ~1.3k | Automated safety check: Pass | MIT |
indranilbanerjee/digital-marketing-pro
Schedule social posts to Twitter/X, Instagram, LinkedIn, TikTok, YouTube, and Pinterest through connected platform MCPs — generating per-platform copy variations, hashtag mixes, media-spec…
minhnv0807/ai-business-skills
A skill your agent uses when the user needs to choose and stand up marketing channels and the tech behind them — channel selection scoring, platform-by-platform setup checklists, business account…
nicepkg/ai-workflow
Plans, creates, and optimizes digital marketing campaigns including content strategy, social media, email marketing, SEO, and AI visibility (GEO).
seranking/seo-skills
Measure AI Search share of voice for a target domain versus competitors across ChatGPT, Perplexity, Gemini, Google AI Overview, and AI Mode.
social-media-skills/skills
Competitor analysis for social media — public-data competitive reconnaissance to find the gap a brand can own.
rongxinzy/RongxinAI
Plan, produce, launch, measure, optimize, and review integrated marketing campaigns.
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.
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.
XBuilderLAB/cheat-on-content
Records that a piece of content went live by writing its URL, platform and publish time into the prediction file header and state file, leaving the prediction text untouched.
XBuilderLAB/cheat-on-content
Develops one content topic at a time through conversation: you bring a theme or experience, the agent digs for angles and writes a draft, with an optional batch brainstorm mode.
XBuilderLAB/cheat-on-content
登记一条视频已拍摄。建 video folder + 询问实际拍摄稿是否与 scripts/<id.md 一致 + buffer +1。与 cheat-publish 配对:拍了进队列,发了出队列。触发词:"拍了"/"拍了 X"/"shot"/"shot it"/"已拍 X"/"录完了"。
XBuilderLAB/cheat-on-content
从配置的热点源(HN / Reddit / YouTube trending / B 站热门 / 等)抓今天的热门话题,去重 + 粗打分 + 写入 candidates.md。绝大部分人没有候选池——这是让"我没素材"问题在 onboarding 第二步就消失的钥匙。触发词:"抓热点"/"fetch trends"/"今天有什么可做的"/"trending now"/"找选题"。
Categories
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. Two modes exist and are never mixed in one run. A full rubric bump (--propose) changes the formula, its dimensions or weights, and must follow a five-step validation protocol with a cross-model review.
Rubric Bump Proposer fits situations like: changing the weights or dimensions of a content scoring formula; recalibrating bucket boundaries after an account has grown; testing a proposed rubric change against past predictions.
Run `npx skills add XBuilderLAB/cheat-on-content --skill cheat-bump -a claude-code`. Or copy the skill folder (skills/cheat-bump in XBuilderLAB/cheat-on-content) into .claude/skills/cheat-bump in your project. Claude Code loads it when a task matches its description.
Run `npx skills add XBuilderLAB/cheat-on-content --skill cheat-bump -a codex`. Or copy the skill folder (skills/cheat-bump in XBuilderLAB/cheat-on-content) into .agents/skills/cheat-bump in your project. Codex loads it when a task matches its description.
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-bump -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-bump, .gemini/skills/cheat-bump, .github/skills/cheat-bump and .opencode/skills/cheat-bump in your project.
SKILL.md names no scripts, command-line tools or credentials: Rubric Bump Proposer is instructions for the agent only. Our summary lists: A cheat-on-content project with rubric_notes.md, a predictions folder and .cheat-state.json; An external LLM chat tool for the cross-model audit. Its frontmatter pre-approves these tools: Bash(*), Read, Write, Edit, Glob, Grep, Skill, Task, mcp__llm-chat__chat.
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.
Our automated static check of SKILL.md found notes only (pre-approves every shell command (allowed-tools: bash)), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.
Rubric Bump Proposer is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 3.5k tokens (SKILL.md is roughly 14k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
Skills that share tags, products or a category with Rubric Bump Proposer: Schedule Social (indranilbanerjee/digital-marketing-pro, 855 stars), 11 Channel Setup Global (minhnv0807/ai-business-skills, 608 stars), Running Marketing Campaigns (nicepkg/ai-workflow, 285 stars) and SEO AI Search Share Of Voice (seranking/seo-skills, 160 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
XBuilderLAB (a GitHub organization) maintains it in XBuilderLAB/cheat-on-content, which has 7,234 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.