mirror of
https://github.com/linshenkx/prompt-optimizer.git
synced 2026-07-01 12:04:21 +08:00
- fold earlier planning notes into a single current-spec and archived history structure - keep manual acceptance steps and real API samples aligned with the refactored analysis/result/compare model - retain supporting workspace notes needed to review version-selection and evaluation behavior changes
41 lines
1.4 KiB
JSON
41 lines
1.4 KiB
JSON
{
|
|
"type": "result",
|
|
"score": {
|
|
"overall": 70,
|
|
"dimensions": [
|
|
{
|
|
"key": "goalAchievement",
|
|
"label": "目标达成度",
|
|
"score": 80
|
|
},
|
|
{
|
|
"key": "outputQuality",
|
|
"label": "输出质量",
|
|
"score": 90
|
|
},
|
|
{
|
|
"key": "constraintCompliance",
|
|
"label": "约束符合度",
|
|
"score": 50
|
|
},
|
|
{
|
|
"key": "promptEffectiveness",
|
|
"label": "提示词引导有效性",
|
|
"score": 60
|
|
}
|
|
]
|
|
},
|
|
"improvements": [
|
|
"当前执行提示词“写一首诗”过于宽泛,未对诗歌风格、主题或格式做出明确约束,导致模型可能自由发挥,无法保证输出一致性。建议补充如‘五言绝句’或‘表达思念之情’等具体要求。",
|
|
"输出结尾的附加语句(——愿这首诗能寄托你心中的思绪)属于未被请求的元评论内容,违反了输出边界,需在提示词中明确禁止此类额外说明。",
|
|
"若希望提高创作稳定性与可预期性,可在提示词中加入示例或参考风格,例如‘请模仿李白风格写一首五言律诗’。"
|
|
],
|
|
"summary": "输出完成任务且质量较高,但提示词缺乏清晰约束,且存在输出边界滑移。",
|
|
"patchPlan": [],
|
|
"metadata": {
|
|
"model": "dashscope",
|
|
"timestamp": 1773732225083,
|
|
"duration": 17528
|
|
}
|
|
}
|