AgentStack
SKILL verified MIT Self-run

Score Optimizer

skill-xiaomoboy-claude-writing-skills-score-optimizer · by xiaomoBoy

Use when the user wants to iterate on a viral-article scoring system itself, calibrate or improve a scoring prompt against labeled samples, or run batch scoring experiments on a fixed article set. Best for prompt-only scoring research where the evaluator scripts stay fixed and only the scoring rubric/prompt is meant to evolve.

No reviews yet
0 installs
6 views
0.0% view→install

Install

$ agentstack add skill-xiaomoboy-claude-writing-skills-score-optimizer

✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

Are you the author of Score Optimizer? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Score Optimizer

这个 skill 只负责一件事:优化文章评分系统本身。

它适合做的是:

  • 调整文章爆款评分 Prompt
  • 批量给样本文集打分
  • 对照标注数据评估评分准确度
  • 研究评分维度、权重、few-shot 示例是否更合理

它不负责:

  • 直接改文章正文
  • 平台改写
  • 发布包装

Working Scope

适用场景:

  • 用户说“优化文章评分 prompt”
  • 用户说“让这个爆款评分系统更准”
  • 用户说“跑一轮样本文集评分实验”
  • 用户说“校准 viral score / composite score 的打分规则”

不适用场景:

  • 用户说“直接把这篇文章改高分”
  • 用户说“按这个评分器去优化文章内容”

后者应改用 article-optimizer

Required Reads

开始前先读:

  1. [references/scoreprogram.md](references/scoreprogram.md)
  2. [references/scoreprompt.md](references/scoreprompt.md)

需要看实现时再读:

  • [scripts/runscoring.py](scripts/runscoring.py)
  • [scripts/evaluate.py](scripts/evaluate.py)
  • [assets/articles/labels.example.json](assets/articles/labels.example.json) —— 标注 schema 示例

样本文集位置(用户自己提供,本 skill 不带样本):

  • assets/articles/samples/ —— 把你自己标注过的 .md 文章放进来,然后 cp labels.example.json labels.json 填上你的评分。详见 [assets/articles/samples/README.md](assets/articles/samples/README.md)

Workflow

  1. 先确认当前任务是在优化评分器,不是在改文章。
  2. 读取 references/score_program.md,按它的实验约束执行。
  3. 默认只改 references/score_prompt.md,不要动评估脚本。
  4. 运行批量评分脚本,再运行评估脚本。
  5. 比较 composite_scoremaespearmanclassification
  6. 记录结果,决定保留还是回滚。

Commands

在 skill 根目录执行:

python3 scripts/run_scoring.py > scoring.log 2>&1
python3 scripts/evaluate.py > eval.log 2>&1
grep "^composite_score:\|^mae:\|^spearman:\|^classification:" eval.log

如果需要图表:

python3 scripts/generate_score_chart.py

Guardrails

  • 默认不要改 scripts/evaluate.py
  • 默认不要改 scripts/run_scoring.py
  • 默认不要安装新依赖
  • 如果只是想提高某一篇文章分数,不要用这个 skill

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet — be the first.

Versions

  • v0.1.0 Imported from the upstream source.