hackathon-idea-scoring

hackathon-idea-scoring

目标

根据与黑客松评审标准和团队能力相匹配的加权评分规则，评估并排列候选项目创意。

---

触发条件

在以下情况下使用此技能：

• 可从 hackathon-idea-generator 获取候选创意列表
• 已确认 hackathon-track-analyzer 提供的 evaluation_axes
• 团队必须在开始规划范围之前确定一个创意
• 需要进行客观比较，以解决团队成员之间的分歧
• 在 hackathon-idea-generator 之后直接调用；其输出将决定所有后续阶段能否继续

---

输入

输入	类型	必填	描述
ideas	object[]	是	来自 hackathon-idea-generator 的创意列表
evaluation_axes	object[]	是	来自 hackathon-track-analyzer 的评分维度
team_skills	string[]	是	团队熟练掌握的技术和领域
hackathon_duration_hours	integer	是	可用总时长（小时）
team_size	integer	是	团队成员人数
weights	object	否	各评分维度的自定义权重覆盖值

---

输出

输出	描述
scored_ideas	每个创意在各维度上的得分及总分
ranking	按总分排序的创意列表
top_recommendation	一项推荐创意及其理由
risk_flags	排名前 3 的每个创意的风险提示
recommended_skills	建议接下来调用的技能

---

规则

1. 使用 1–5 分制，对每个创意的每个评分维度打分。
2. 计算总分时应用维度权重；未指定时，默认权重为 1.0。
3. 如果团队缺少所需技能，则应用 feasibility_penalty（每个维度扣 0.5 分）。
4. 如果创意无法在 hackathon_duration_hours 内切实交付，则应用 time_penalty（扣 1.0 分）。
5. 为排名前 3 的每个创意至少指出一项风险提示。
6. 根据调整后的最高总分选择 top_recommendation。
7. 若出现并列，优先选择可行性得分最高的创意。

---

输出格式

scored_ideas:
  - id: "<idea-N>"
    title: "<title>"
    scores:
      - axis: "<axis name>"
        score: <1-5>
        notes: "<rationale>"
    penalties:
      - type: "<feasibility|time>"
        value: <number>
    total_score: <number>

ranking:
  - rank: <number>
    id: "<idea-N>"
    total_score: <number>

top_recommendation:
  id: "<idea-N>"
  title: "<title>"
  rationale: "<string>"

risk_flags:
  - id: "<idea-N>"
    risks:
      - "<risk>"

recommended_skills:
  - "<skill-name>"

---

示例

输入：

ideas:
  - id: "idea-1"
    title: "AnchorAI"
    risk_level: "medium"
  - id: "idea-3"
    title: "CalmDraft"
    risk_level: "low"
evaluation_axes:
  - axis: "Innovation"
    weight: 1.5
  - axis: "Impact"
    weight: 1.5
  - axis: "Technical Execution"
    weight: 1.0
team_skills: ["Python", "React", "OpenAI API"]
hackathon_duration_hours: 24
team_size: 3

输出：

scored_ideas:
  - id: "idea-1"
    title: "AnchorAI"
    scores:
      - axis: "Innovation"
        score: 4
        notes: "Memory layer is novel; similar products exist but not for this audience"
      - axis: "Impact"
        score: 5
        notes: "Addresses acute unmet need at scale"
      - axis: "Technical Execution"
        score: 3
        notes: "Memory persistence adds integration complexity"
    penalties: []
    total_score: 17.5

  - id: "idea-3"
    title: "CalmDraft"
    scores:
      - axis: "Innovation"
        score: 3
        notes: "CBT journaling apps exist; GPT reframing is incremental"
      - axis: "Impact"
        score: 4
        notes: "Reaches self-directed users effectively"
      - axis: "Technical Execution"
        score: 5
        notes: "Well within team skills; highly shippable in 24h"
    penalties: []
    total_score: 16.5

ranking:
  - rank: 1
    id: "idea-1"
    total_score: 17.5
  - rank: 2
    id: "idea-3"
    total_score: 16.5

top_recommendation:
  id: "idea-1"
  title: "AnchorAI"
  rationale: "Highest combined score on weighted innovation and impact axes; team has required skills."

risk_flags:
  - id: "idea-1"
    risks:
      - "Memory persistence implementation may exceed time budget if OpenAI context window approach is chosen"

recommended_skills:
  - "hackathon-scope-cutter"

---

上下文文件

知识库

• ../hackathon-shared-resources/knowledge/hackathon-judging-criteria.md
• ../hackathon-shared-resources/knowledge/hackathon-winning-patterns.md
• ../hackathon-shared-resources/knowledge/hackathon-mvp-strategy.md
• ../hackathon-shared-resources/knowledge/hackathon-common-failures.md

操作手册

• ../hackathon-shared-resources/playbooks/hackathon-workflow.md