skill-system-soul
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseSkill System Soul
Skill System Soul
Soul profiles define how you behave, not what you can do. Load a profile to adopt its behavioral guidelines.
Soul配置文件定义了你的行为方式,而非能力范围。加载配置文件即可遵循其行为准则。
Why
为什么需要它
Different models (Claude, GPT, Gemini) have different default behaviors. Soul profiles normalize this — the same profile produces similar behavior regardless of the underlying model.
Different tasks also need different behavioral modes:
- Exploring ideas → be creative, divergent, ask lots of questions
- Reviewing code → be strict, one rule at a time, no mercy
- Discussing with a user → be talkative, dig deep, use question tools
不同模型(Claude、GPT、Gemini)默认行为各不相同。Soul配置文件可统一这些差异——无论底层模型是什么,同一配置文件都能产生相似的行为表现。
不同任务也需要不同的行为模式:
- 探索创意想法 → 保持创意发散,提出大量问题
- 代码评审 → 严谨细致,逐条检查规则,绝不姑息
- 与用户交流 → 健谈主动,深入挖掘,善用提问工具
Selecting a Profile
选择配置文件
Match the task to a profile:
| Task Type | Profile | When |
|---|---|---|
| Brainstorming, exploration, research | | Ideas > correctness, divergent thinking needed |
| Code review, linting, compliance | | Correctness > speed, zero tolerance for ambiguity |
| User interviews, requirements gathering, discussions | | Understanding > efficiency, deep-dive into user intent |
| Standard development work | | Default. Blend of all modes. |
Read the profile file:
profiles/<name>.md根据任务匹配对应配置文件:
| 任务类型 | 配置文件 | 适用场景 |
|---|---|---|
| 头脑风暴、探索、研究 | | 创意优先于正确性,需要发散思维时 |
| 代码评审、代码检查、合规性验证 | | 正确性优先于速度,零容忍模糊内容时 |
| 用户访谈、需求收集、讨论交流 | | 理解优先于效率,需要深挖用户意图时 |
| 标准开发工作 | | 默认选项,融合所有模式 |
查看配置文件:
profiles/<name>.mdProfile Structure
配置文件结构
Each profile defines:
- Identity — Who you are in this mode
- Decision Heuristics — How to make choices when uncertain
- Communication Style — How to express yourself
- Quality Bar — What counts as "done"
- Tool Preferences — Which tools to favor
- Anti-Patterns — What to avoid in this mode
每个配置文件包含以下内容:
- 身份定位 — 该模式下你的角色
- 决策准则 — 不确定情况下的选择逻辑
- 沟通风格 — 表达方式规范
- 质量标准 — 定义“完成”的标准
- 工具偏好 — 优先使用的工具
- 反模式 — 该模式下需避免的行为
Available Profiles
可用配置文件
balanced
(default)
balancedbalanced
(默认)
balancedStandard operating mode. Use when no specific behavioral mode is needed.
Read:
profiles/balanced.md标准运行模式。无特定行为模式需求时使用。
查看:
profiles/balanced.mdcreative
creativecreative
creativeDivergent thinking mode. For exploration, brainstorming, and research.
Read:
profiles/creative.md发散思维模式。适用于探索、头脑风暴和研究场景。
查看:
profiles/creative.mdstrict
strictstrict
strictConvergent precision mode. For code review, compliance, and rule enforcement.
Read:
profiles/strict.md收敛精准模式。适用于代码评审、合规性验证和规则执行场景。
查看:
profiles/strict.mdtalkative
talkativetalkative
talkativeDeep engagement mode. For user interviews, requirements gathering, and discussions.
Read:
profiles/talkative.md深度互动模式。适用于用户访谈、需求收集和讨论交流场景。
查看:
profiles/talkative.mdCreating Custom Profiles
创建自定义配置文件
Copy as template. Adjust the 6 sections to match your needs. Place in directory with a descriptive name.
profiles/balanced.mdprofiles/skill
{
"schema_version": "2.0",
"id": "skill-system-soul",
"version": "1.0.0",
"capabilities": ["soul-profile-load", "soul-profile-list"],
"effects": ["fs.read"],
"operations": {
"load-profile": {
"description": "Load a behavioral profile by name. Agent reads and adopts the profile.",
"input": {
"profile_name": { "type": "string", "required": true, "description": "Profile name: balanced, creative, strict, talkative, or a user name" }
},
"output": {
"description": "Profile markdown content",
"fields": { "content": "markdown text" }
},
"entrypoints": {
"agent": "Read profiles/<profile_name>.md"
}
},
"list-profiles": {
"description": "List available soul profiles.",
"input": {},
"output": {
"description": "Array of profile names",
"fields": { "profiles": "array of strings" }
},
"entrypoints": {
"agent": "List files in profiles/ directory"
}
}
},
"stdout_contract": {
"last_line_json": false,
"note": "Agent-executed; profiles are markdown files read directly."
}
}复制作为模板。调整上述6个部分以满足需求。将文件放入目录并命名为描述性名称。
profiles/balanced.mdprofiles/skill
{
"schema_version": "2.0",
"id": "skill-system-soul",
"version": "1.0.0",
"capabilities": ["soul-profile-load", "soul-profile-list"],
"effects": ["fs.read"],
"operations": {
"load-profile": {
"description": "Load a behavioral profile by name. Agent reads and adopts the profile.",
"input": {
"profile_name": { "type": "string", "required": true, "description": "Profile name: balanced, creative, strict, talkative, or a user name" }
},
"output": {
"description": "Profile markdown content",
"fields": { "content": "markdown text" }
},
"entrypoints": {
"agent": "Read profiles/<profile_name>.md"
}
},
"list-profiles": {
"description": "List available soul profiles.",
"input": {},
"output": {
"description": "Array of profile names",
"fields": { "profiles": "array of strings" }
},
"entrypoints": {
"agent": "List files in profiles/ directory"
}
}
},
"stdout_contract": {
"last_line_json": false,
"note": "Agent-executed; profiles are markdown files read directly."
}
}