赛事评审打分器

技能文档

---
name: judging-rubric-grader
slug: judging-rubric-grader
displayName: 赛事评审打分器
version: 1.1.1
category: ai-agent
display_name: 赛事评审打分器
title: 赛事评审打分器
author: 诺律(SynomosAI)
author_note: '"SynomosAI" — 未申请实体注册、未申请商标注册（not a registered legal entity; no trademark registered）'
license: MIT
allowed-tools: Read, Glob, Grep, Bash, Write, WebFetch
license_scope: 代码（.py 文件）MIT；本 SKILL.md 与其中理论文本不在 MIT 覆盖范围内（见文末「License & attribution」）
description: 赛事评审打分器——基于 UIBC 赛事规范的零依赖决策支持工具，JSON IR 输出，覆盖TH-EVT-003域。 适用场景：当需要赛事评审打分器能力时使用本技能；触发词：赛事、判定器、TH-EVT-003、judging-rubric-grader。
tags: [赛事, 判定器, TH-EVT-003, judging-rubric-grader, 评审打分]
description_zh: "赛事评审打分器——基于 UIBC 赛事规范的零依赖决策支持工具，JSON IR 输出，覆盖TH-EVT-003域。"
description_en: "Contest judging rubric scorer: zero-dependency decision support with JSON IR output (UIBC spec, domain TH-EVT-003)."
classification:
  skillhub: ["community-events"]
  iso_25010: ["Usability"]
risk_tier: medium
risk_rationale: "门禁批量补丁默认 medium（按默认策略设计为中等风险）：调用 API/读写文件/本地执行，最小权限，密钥走 Secret；若实际涉对外代执行(push/发布/发送/上架)须人工复评，改 high 并走人工批准。"
platforms: [windows, macos, linux]
read_when:
  - 赛事评审要严格按 UIBC 规范客观、标准化打分
  - 评分维度多/权重复杂，要减少人为判断偏差
  - 要输出符合 TH-EVT-003 的 JSON IR 对接后续处理或系统接口
  - 用户说「帮我按评分表算总分和等级」
ai_protocol: true
spec: usk/1.0
interface:
  type: cli
  entry_point: judging-rubric-grader.py
  runtime: python3
  call_pattern: args
input_schema:
  args:
    type: string
    description: "命令行参数透传给主脚本，具体取值见各包 --help"
    required: false
output_schema:
  stdout:
    type: string
    description: "标准输出为 JSON 结果，含逐项检查与总体结论"
  exit_code:
    type: integer
    description: "0=PASS/完成；1=FAIL/有警告或待补强；2=用法或参数错误"
capabilities:
  - cli_tool
permissions:
  network: false
  filesystem: false
  subprocess: false
requirements:
  python_packages: []
  min_python: "3.8"
changelog: "1.1.1: Bilingual rewrite (full EN+ZH, English first), added read_when; 1.0.0: initial upload (USK v3)"

---

> 凡自治之物 · LGD 治理学 · 创造术家族 ｜ 评测协议：见平台质量门禁规范

## English

### When to use
This skill applies when contest entries must be scored objectively and standardly under the UIBC rules — especially when the rubric is complex and human judgment may drift. It generates a zero-dependency decision-support report. It is also the preferred choice when the task must emit TH-EVT-003-domain JSON IR data for downstream processing or system integration.

### Usage
This is a document-style skill with no local script to run; invoke it with a structured instruction. Provide the entrant ID, contest event type and the corresponding UIBC scoring criteria. You may pin a specific domain rule version (e.g. TH-EVT-003) to lock the scoring logic. The skill returns a standardized JSON IR block that can be embedded directly into a workflow or log system without extra parsing.

### Workflow
1. Initialize: the caller prepares the entrant data and the chosen UIBC rule version.
2. Validate: the system checks that the input is compatible with the TH-EVT-003 domain.
3. Compute: the built-in scoring algorithm weights and scores each dimension.
4. Generate: build the raw output block conforming to the JSON IR standard.
5. Verify: check that the output contains the required metadata and error flags.
6. Deliver: return the final scoring report to the caller for archiving or display.

### Evidence & safety statement
- Evidence: TBD (no evaluation / badge link yet; update here once it is produced)
- Version: 1.1.1
- Security: no independent security scan has been performed (TBD); the conclusion is pending and no endorsement is given until it is completed.

## 中文

### 何时用
本技能适用于需要严格依据 UIBC 赛事规范进行客观、标准化的赛事打分场景。特别是在评分标准复杂，存在人为判断偏差风险时，调用该技能生成零依赖的决策支持报告。此外，当任务需要输出符合 TH-EVT-003 域的 JSON IR 格式数据以对接后续数据处理或系统接口时，优先使用此技能。

### 用法
由于本技能为文档型技能，通过输入结构化指令进行调用。需提供包含选手 ID、竞赛项目类型及对应 UIBC 评分细则的输入参数。用户可指定特定域规则版本（如 TH-EVT-003 规范）以锁定评分逻辑。最终将返回标准化 JSON IR 数据块，无需额外解析即可直接嵌入工作流或日志系统中。

### 流程
1. 初始化：调用者准备待评审的选手数据及指定的 UIBC 赛事规则版本信息。
2. 校验：系统自动核查输入内容是否兼容 TH-EVT-003 域要求。
3. 计算：基于内置的评分算法对各项指标进行加权运算与打分。
4. 生成：构建符合 JSON IR 标准格式的原始数据输出块。
5. 验证：检查输出结果中是否包含必要的元数据及错误标记。
6. 交付：将最终评分报告返回给调用端以供归档或展示。

### 证据与安全声明
- 证据：TBD（暂无评测/徽章链接；产生后在此更新实际链接）
- 版本：1.1.1
- 安全：未做独立安全扫描（TBD），结论待补；补全前不作推荐背书。

使用说明

# judging-rubric-grader · 赛事评审打分器

[![LGD](https://img.shields.io/badge/LGD-Governance-blue)](https://github.com/zhaoxinghua09-cell/lgd-theory)

> AI-assisted decision-support CLI for UIBC contest operations. NOT final adjudication — verify with the organizing committee. Text verified: 2026-09-26.

## English

### What it does
Grades submissions by **multi-dimension rubric weighting** per the **UIBC judging spec (weighted rubric)** (rule source pending manual verification; the latest UIBC charter prevails).

- **Insufficient input → strict presumption**: a dimension missing from `scores` is counted as **0** (lowering the total) and stated in `warnings` (`rc=1`) — never silently skipped or full-marked. A weight sum ≠ 1 also warns.
- **Out of scope**: judge deliberation, appeals, overall endorsement of submission safety/correctness.

### Install & run (zero-dependency, Python 3.8+)
```bash
python judging-rubric-grader.py --criteria '[{"dim":"innovation","weight":0.4},{"dim":"engineering","weight":0.3},{"dim":"presentation","weight":0.3}]' --scores '{"innovation":90,"engineering":85,"presentation":80}'   # complete → rc=0
python judging-rubric-grader.py --criteria '[{"dim":"innovation","weight":0.5},{"dim":"compliance","weight":0.5}]' --scores '{"innovation":90}'                                                   # missing dim → strict 0, rc=1
python judging-rubric-grader.py                                                                                                                                                  # missing args → rc=2
python judging-rubric-grader.py --demo                                                                                                                                           # 3 built-in smoke cases
```
Params: `criteria` (JSON dim/weight list, required) / `scores` (JSON dim→score object, required) / `pass_line` (default 60).

### Output
Deterministic JSON IR with `rc` (**0** = ok / **1** = completed with warnings / **2** = insufficient or invalid input), `total`, `grade`, `per_dimension`, `warnings`, and `aigc_mark` (GB 45438-2025 metadata; `is_generated: true`, label basis noted as pending verification).

### Rule-source review checklist (pending manual verification)
| Rule | Used for | Status |
|---|---|---|
| UIBC judging spec · rubric | weighted total & grade mapping | pending manual check |
| Grade breakpoints 90/80/70 / pass line | grade mapping values | pending manual check |

### Compliance & IP
- LGD-Powered badge · theory TH-EVT-003 · ambassador 诺律 (LAW-domain lead voice)
- Decision support only; human review required.
- **Layered licence**: code `judging-rubric-grader.py` = MIT; this `README.md` / `SKILL.md` and all theory text are **not covered by MIT** (all rights reserved).

### Family set
| Contest stage | Tool | Ambassador |
|---|---|---|
| Registration / submission portal | contest-submission-portal | 诺声 |
| Problem-set generation | problem-set-generator | 诺声 |
| Judging & scoring | judging-rubric-grader | 诺律 |
| Anti-cheat / identity check | anti-cheat-identity | 诺卫 |
| Plagiarism check | originality-check | 诺源 |

## 中文

### 它做什么
依据 **UIBC 评审规范（rubric 加权）**（规则源待人工核对；以最新 UIBC 章程为准）对参赛作品做**多维度加权打分**。

- **输入不足 → 从严推定**：`scores` 中缺的维度按 **0 分**计（拉低总分），并在 `warnings` 中说明（`rc=1`）——绝不静默跳过或给满分；权重之和 ≠ 1 也告警。
- **不做什么**：评委合议、申诉处理、对作品安全性/正确性的总体背书。

### 安装与运行（零依赖，Python 3.8+）
```bash
python judging-rubric-grader.py --criteria '[{"dim":"创新","weight":0.4},{"dim":"工程","weight":0.3},{"dim":"表达","weight":0.3}]' --scores '{"创新":90,"工程":85,"表达":80}'   # 参数齐 → rc=0
python judging-rubric-grader.py --criteria '[{"dim":"创新","weight":0.5},{"dim":"合规","weight":0.5}]' --scores '{"创新":90}'                                                   # 缺维度 → 从严 0 分，rc=1
python judging-rubric-grader.py                                                                                                                                                  # 缺参数 → rc=2
python judging-rubric-grader.py --demo                                                                                                                                           # 3 个内置冒烟用例
```
参数：`criteria`（JSON 维度/权重列表，必填）/ `scores`（JSON 维度→分数对象，必填）/ `pass_line`（默认 60）。

### 输出
确定性 JSON IR，含 `rc`（**0**=ok / **1**=完成有告警 / **2**=输入不全或非法）、`total`、`grade`、`per_dimension`、`warnings` 与 `aigc_mark`（GB 45438-2025 元数据；`is_generated: true`，标注依据待核对）。

### 规则源待人工核对清单
| 规则 | 用途 | 状态 |
|---|---|---|
| UIBC 评审规范·rubric | 加权总分与等级映射 | 待人工核对 |
| 等级断点 90/80/70/达标线 | 等级映射取值 | 待人工核对 |

### 合规与署名
- LGD-Powered 徽章 · 理论 TH-EVT-003 · 大使 诺律（LAW 域首席发声人）
- 本工具为决策支持，须人工复核。
- **分层许可**：代码 `judging-rubric-grader.py` = MIT；本 `README.md` / `SKILL.md` 及一切理论文本**不在 MIT 覆盖范围内**（保留所有权利）。

### 家族合集
| 赛事环节 | 工具 | 大使 |
|---|---|---|
| 报名/提交门户 | contest-submission-portal | 诺声 |
| 赛题生成 | problem-set-generator | 诺声 |
| 评审打分 | judging-rubric-grader | 诺律 |
| 反作弊/身份核验 | anti-cheat-identity | 诺卫 |
| 作品查重 | originality-check | 诺源 |

---

## License & attribution
**分层许可（务必并读；代码许可不覆盖文档与理论文本）**

- **代码**（本包 `.py` 文件）：**MIT**
- **本 `README.md` / `SKILL.md` 及其中的理论文本、方法论与一切理论表述**：**不在 MIT 覆盖范围内**，保留所有权利（All rights reserved）

```
© 2026 SynomosAI · China. All rights reserved.
理论署名 (attribution) : LGD（Lifecycle Governance Doctrine / 全程治理论）— SynomosAI initiative
名称状态 (name status)  : "SynomosAI" — 未申请实体注册、未申请商标注册
                        (not a registered legal entity; no trademark registered)
生产参考部署 (production reference, self-reported) : 无（未声明外部生产参考部署）
                    ← 非认证、非背书、非监管认可（not a certification or endorsement）
代码许可 (code license) : 本包未附 LICENSE 文件（许可待定）
                    本文本与理论表述不在任何代码许可覆盖范围内
引用格式 (cite as)      : 本资产无 DOI
```
（上列为《LGD 对外表述规范》§3.2 统一块，**整体复制、未删改**；发布/重提前须照最新版本复核。）

如何安装此技能？

访问技能市场，点击「安装」按钮，按提示将技能包放入 AI 编程助手的 skills 目录即可。