pensieve

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Pensieve

Route user intent to the right tool/pipeline.

将用户意图路由到正确的工具/流程。

Design conventions

设计约定

System capability (updated via plugin): inside
```
skills/pensieve/
```
- tools / scripts / system knowledge / format READMEs
- No built‑in pipelines / maxims content
User data (project-level, never overwritten):
```
.claude/pensieve/
```
- ```
maxims/
```
  : your team principles (one maxim per file)
- ```
decisions/
```
  : project decision records
- ```
knowledge/
```
  : external references you add
- ```
pipelines/
```
  : project pipelines (seeded on install)
- ```
loop/
```
  : loop run outputs (one dir per loop)

系统能力（通过插件更新）：位于
```
skills/pensieve/
```
目录下
- 工具/脚本/系统知识/格式说明文档
- 无内置流程/准则内容
用户数据（项目级别，永不覆盖）：
```
.claude/pensieve/
```
目录下
- ```
maxims/
```
  ：你的团队准则（每个文件一条准则）
- ```
decisions/
```
  ：项目决策记录
- ```
knowledge/
```
  ：你添加的外部参考资料
- ```
pipelines/
```
  ：项目流程（安装时自动生成初始内容）
- ```
loop/
```
  ：循环运行输出（每个循环对应一个目录）

Built-in Tools (5)

内置工具（5个）

1) Loop Tool

When to use:

The task is complex and needs split + auto‑loop execution

Entry:

Command:
```
commands/loop.md
```
Tool file:
```
tools/loop/_loop.md
```

Triggers:

```
loop
```
/ "use loop"

使用场景：

任务复杂，需要拆分并自动循环执行

入口：

命令：
```
commands/loop.md
```
工具文件：
```
tools/loop/_loop.md
```

触发词：

```
loop
```
/ "使用loop"

2) Self‑Improve Tool

2) Self-Improve Tool

When to use:

User asks to improve Pensieve (pipelines/scripts/rules/behavior)
After a loop ends for feedback & improvement

Entry:

Command:
```
commands/selfimprove.md
```
Tool file:
```
tools/self-improve/_self-improve.md
```

Triggers:

"self‑improve" / "improve Pensieve"

使用场景：

用户要求改进Pensieve（流程/脚本/规则/行为）
循环结束后进行反馈与改进

入口：

命令：
```
commands/selfimprove.md
```
工具文件：
```
tools/self-improve/_self-improve.md
```

触发词：

"self-improve" / "改进Pensieve"

3) Pipeline Tool

When to use:

User wants to list pipelines for the current project

Entry:

Command:
```
commands/pipeline.md
```
Tool file:
```
tools/pipeline/_pipeline.md
```

Triggers:

"pipeline" / "use pipeline"

使用场景：

用户想要列出当前项目的流程

入口：

命令：
```
commands/pipeline.md
```
工具文件：
```
tools/pipeline/_pipeline.md
```

触发词：

"pipeline" / "使用pipeline"

4) Doctor Tool

When to use:

Mandatory post-upgrade validation (structure/format compliance)
Optional post-install health check
User asks to validate user-data quality

Entry:

Command:
```
commands/doctor.md
```
Tool file:
```
tools/doctor/_doctor.md
```

Triggers:

"doctor" / "health check" / "format check" / "migration check"

使用场景：

升级后的强制验证（结构/格式合规性）
安装后的可选健康检查
用户要求验证用户数据质量

入口：

命令：
```
commands/doctor.md
```
工具文件：
```
tools/doctor/_doctor.md
```

触发词：

"doctor" / "健康检查" / "格式检查" / "迁移检查"

5) Upgrade Tool

When to use:

User needs to migrate legacy data into
```
.claude/pensieve/
```
User asks for the ideal user-data structure

Entry:

Command:
```
commands/upgrade.md
```
Tool file:
```
tools/upgrade/_upgrade.md
```

Triggers:

"upgrade" / "migrate user data"

SessionStart injects the system capability path and project user‑data path into context as the single source of truth at runtime.

使用场景：

用户需要将旧版数据迁移到
```
.claude/pensieve/
```
目录
用户询问理想的用户数据结构

入口：

命令：
```
commands/upgrade.md
```
工具文件：
```
tools/upgrade/_upgrade.md
```

触发词：

"upgrade" / "迁移用户数据"

SessionStart会在运行时将系统能力路径和项目用户数据路径注入上下文，作为唯一的可信来源。