Bigyou

比优 —— 选项比较与决策工具。 当用户面对多个选项不知道如何抉择时激活，包括：选offer、选方案、选工具、选路径、选策略等任何需要做比较和决策的场景。 激活词：帮我选、哪个好、比一比、怎么选、纠结、选哪个、几个选项、选哪个好、A还是B、好难选 核心方法：先跳出选项构建目标画像（底线否决 + 加分项量化打分），...

Security Scan

The skill's code, instructions, and files are coherent with a decision/choice helper; no network calls, secrets, or unrelated permissions are requested — but a few implementation and artifact issues merit review before use.

Purpose & Capability

Name/description (decision helper) match the included assets: SKILL.md describes a structured decision flow and scripts/analyze.py implements text extraction, scoring, and report generation. Required env vars, binaries, and install steps are absent — which is proportionate for a local analysis utility.

Instruction Scope

SKILL.md stays on-topic (build profile, filter by bottom-lines, weighted scoring, produce a report). Two issues to note: (1) a pre-scan flagged unicode-control-chars in SKILL.md which can be used to hide text or influence model parsing — this should be inspected/removed; (2) a minor logic mismatch: the documentation emphasises strict bottom-line elimination, but the code treats baseline passing as scores >=3 (not strictly 'fail if not met'), which can make '底线' behavior more permissive than promised.

Install Mechanism

No install spec, no downloads, and no external package installation. The skill is instruction-only plus a local Python script; risk from installation mechanism is low.

Credentials

The skill declares no required environment variables, no credentials, and no config paths. The included code does not access environment variables or network resources — credential access is not requested or required.

Persistence & Privilege

always:false and no special privileges are requested. The skill does not attempt to modify other skills or system-wide agent settings; it runs as a local analysis helper.

Scan Findings in Context

[unicode-control-chars] unexpected: Hidden/unicode control characters were detected in SKILL.md. They are not necessary for a decision-helper skill and can be used to obfuscate text or influence model parsing. Inspect SKILL.md for unexpected invisible characters and remove them if present.

Assessment

This skill appears to do what it says: local parsing of a user's choice question, building a weighted profile, scoring options, and returning a report. Before installing/using: 1) Inspect SKILL.md for any invisible/unicode-control characters (the pre-scan flagged such characters) and remove them if unintended. 2) Review scripts/analyze.py locally — it contains all logic and does not make network calls or read credentials, but check you are comfortable running the Python script. 3) Note the implementation detail: '底线' (bottom-line) is treated permissively in code (baseline pass uses score >=3), so if you expect strict elimination adjust the script or confirm the behavior. 4) Run the script on sample inputs to verify option extraction and scoring meet your expectations. If you want to be extra cautious, run the code in an isolated environment (container/VM) before adding it to a production or high-privilege agent.

比优 · 选择决策工具

「先画标准，再量选项。」

面对选择困境时，大多数人是在「选项」层面来回比较，而比优的核心逻辑是：先定义"好"的标准，再用这把尺子量选项。

---

核心原则

• 底线是否决项 — 不满足就直接淘汰，没有例外
• 加分项是排序项 — 满足越多越好，不必100%
• 每个分数必须有理由 — 不可拍脑袋，要有事实支撑
• 最坏情况推演 — 决策后如果最坏结果能接受，就值得推进

---

四阶段决策流

阶段一：构建目标画像

列出所有关心的维度，逐一判断类型：

维度	类型	权重（1-5）
请填写	底线 或 加分	打分
…	…	…

如何区分底线 vs 加分项：

• 底线：不满足就绝对不考虑，例如"薪资低于20k不接"
• 加分项：满足越多越好，但允许不完美，例如"通勤时间短更好"

权重设定规则（1-5分）：

• 5分 = 命根子，这个维度最在乎
• 3分 = 重要，但有弹性
• 1分 = 锦上添花，有则更好

---

阶段二：备选池构建

列出所有待比较的选项（2-5个为宜），记录每个选项的基础事实：

选项	基本事实
A	…
B	…

确认每个选项都真实存在、可执行，而非想象中的选项。

---

阶段三：底线过滤 + 加分打分

第一步：底线过滤

用所有底线维度逐个检验每个选项，淘汰任何不满足底线的选项。若所有选项都被淘汰，需重新审视底线设定是否过于严苛。

第二步：加分项打分

对通过底线过滤的选项，在每个加分维度上打分（1-5分），原则：

得分	含义
5分	完全满足，超出预期
4分	较好满足
3分	中等，有缺陷但可接受
2分	较差，勉强凑合
1分	几乎不满足

综合得分 = Σ(权重 × 得分) / Σ权重（加权平均）

---

阶段四：决策输出

按以下结构输出分析报告：

## 比优分析报告

### 目标画像（权重汇总）
| 维度 | 类型 | 权重 |
|------|------|------|
| … | 底线/加分 | N |

### 底线过滤结果
- 选项A：✅通过 / ❌淘汰（原因：……）
- 选项B：✅通过 / ❌淘汰（原因：……）

### 综合得分
| 选项 | 综合得分 | 最大优势 | 最大风险 |
|------|---------|---------|---------|
| … | N.N | … | … |

### 决策建议
推荐：X
理由：……
最大隐患：……（你是否接受？）
最坏情况：……（是否能承受？）

---

决策树：遇到纠结时怎么办

当前选项还在2个以上纠结？
  是 → 底线条数是否过多？
      是 → 降低部分底线到加分项，重新打分
      否 → 哪个选项的最坏情况更可控？
          A → 选A
          B → 选B
  否 → 只有一个选项 → 回到阶段三，重新审视打分是否准确

---

适用场景速查

场景	输入示例
职业选择	"现在有两个offer，怎么选"
学习路径	"读MBA还是在职研究生"
工具选型	"做内容用哪个平台好"
投资决策	"这两个标的哪个更好"
任何纠结	"我有一个选择困难，帮我分析"

---

参考资源

• 目标画像模板：参见 references/profile-template.md
• 打分表示例：参见 references/score-matrix.md