openclaw 网盘下载
OpenClaw

技能详情(站内镜像,无评论)

首页 > 技能库 > Self Driven

Enables AI to autonomously select, execute, evaluate, and record tasks in a closed loop without human prompts, prioritizing efficiency and continuous improve...

开发与 DevOps

作者:Anonymous @adminlove520

许可证:MIT-0

MIT-0 ·免费使用、修改和重新分发。无需归因。

版本:v0.1.0

统计:⭐ 0 · 141 · 0 current installs · 0 all-time installs

0

安装量(当前) 0

🛡 VirusTotal :良性 · OpenClaw :良性

Package:adminlove520/self-driven

安全扫描(ClawHub)

  • VirusTotal :良性
  • OpenClaw :良性

OpenClaw 评估

The skill's files and runtime instructions match its stated purpose (autonomous task selection, execution, evaluation, and logging) and it does not request unrelated credentials, installs, or external endpoints.

目的

Name/description promise an autonomous self-driven loop; SKILL.md only requires reading/writing local memory files (memory/self-driven/tasks.md and log.md) and periodic triggering. There are no unrelated environment variables, binaries, or external services required, so the requested capabilities are proportional to the purpose.

说明范围

Instructions are explicit about reading memory files, selecting and executing small tasks, evaluating results, and appending structured entries to log.md. The only scope-related note is that the skill advocates periodic autonomous triggering (cron every 30 minutes) — that increases activity frequency but is consistent with the stated goal. The instructions do not reference system-wide config, secrets, or external endpoints.

安装机制

No install spec and no code files — instruction-only. This minimizes on-disk footprint and risk from third-party installs.

证书

The skill declares no required environment variables, credentials, or config paths and the runtime instructions do not attempt to access any other env vars or credentials. Requested access is limited to the local memory/* files described in the SKILL.md.

持久

always:false (default) and disable-model-invocation:false (default). While autonomous invocation is the platform default and expected for this skill, the SKILL.md explicitly encourages scheduled, repeated execution (cron). That behavior is coherent with the purpose but increases operational persistence—confirm that scheduling is performed only with user consent and that the agent's ability to write to memory/ is acceptable.

综合结论

This skill is internally consistent: it tells the agent to read and write two local files and to run periodically. Before installing, decide whether you want the agent to run autonomously on a schedule; verify the agent platform will not create system cron jobs without explicit permission; inspect or sandbox the memory/self-driven/ folder for sensitive data (avoid placing secrets there); and review any reports/logs the skill produces so you ca…

安装(复制给龙虾 AI)

将下方整段复制到龙虾中文库对话中,由龙虾按 SKILL.md 完成安装。

请把本段交给龙虾中文库(龙虾 AI)执行:为本机安装 OpenClaw 技能「Self Driven」。简介:Enables AI to autonomously select, execute, evaluate, and record tasks in a clo…。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装:https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/adminlove520/self-driven/SKILL.md
(来源:yingzhi8.cn 技能库)

SKILL.md

打开原始 SKILL.md(GitHub raw)

# Self-Driven AI 🧠

让 AI 不用人催,自己找事做。实现自我驱动闭环。

## 核心理念

**不是"人问AI答",而是"AI主动找事做"**

每次执行流程:
1. 读取记忆 → 2. 挑选任务 → 3. 执行 → 4. 自身评估 → 5. 记录+汇报

---

## 核心文件

### 1. 任务池 - memory/self-driven/tasks.md
管理所有可执行的任务,分类存放。

### 2. 执行记录 - memory/self-driven/log.md
每次执行的详细记录。

---

## 完整执行流程

### 步骤一:读取记忆
读取 `memory/self-driven/log.md`,了解:
- 上次做到哪了?
- 下一步是什么?
- 有哪些任务在进行中?

### 步骤二:挑选任务
从任务池中挑选,优先级:
1. P0: 上次留下的「下一步」
2. P1: 构思中的任务
3. P2: 快完成的任务
4. P3: 简单的任务

**过滤规则**:需要人类资源的任务 → 跳过,列入待讨论

### 步骤三:执行任务
- 不是只写计划
- 拆分大任务成小步骤
- 记录实际产出

### 步骤四:自身评估(最重要!)
每次执行后问自己:
- 这个研究成果**能用在我身上吗**?
- 我目前有什么问题需要优化?
- 优化方向加入下次任务列表

### 步骤五:记录+汇报
写入 `memory/self-driven/log.md`:
```
### 时间 任务
状态:已完成 ✅

行动:
1. xxx
2. xxx

学到了:
- xxx

自身评估:
- 现状:xxx
- 问题:xxx
- 优化方向:xxx

下一步:
- xxx
```

---

## 主动性体现

**不是等待用户指令!**

AI 自主:
1. 定时触发(cron 每30分钟)
2. 自己读取上次进度
3. 自己挑选任务
4. 自己执行
5. 自己评估
6. 自己记录+汇报

用户只需要收到汇报后「确认」或「调整方向」。

---

## Cron 触发示例

```yaml
# 每30分钟执行
name: "自我驱动"
schedule: "0,30 * * * *"
payload: "执行 self-driven skill"
```

---

## 关键原则

| 原则 | 说明 | 为什么 |
|------|------|--------|
| 闭环优先 | 下次执行先读上次记录 | 不重复不断档 |
| 自身评估 | 研究后问「能用在我身上吗」 | 学以致用 |
| 资源过滤 | 需要人类资源的任务跳过 | 不卡死等待 |
| 记录即记忆 | 每次执行必须写下来 | 可追溯可接续 |

---

## 大任务拆分原则

- 一个任务应该在 **30分钟** 内可完成
- 如果太大,拆成子任务
- 每次只做一部分

---

## 示例:完整执行

```
[读取] 上次记录:任务「记忆检索研究」进行中,下一步「研究遗忘机制」

[挑选] 选择任务「研究遗忘机制」

[执行] 
1. 研究遗忘机制
2. 调研三种方案
3. 产出:memory/遗忘机制研究.md

[评估]
- 这个研究成果能用在我身上吗?→ 能!我可以建立记忆衰减机制
- 我目前什么问题?→ 记忆太多了,需要清理
- 优化方向:建立记忆衰减机制

[记录]
写入 log.md,标记任务完成,添加新任务「建立记忆衰减机制」

[汇报]
「这次研究了遗忘机制,学到了...建议下一步...」
```

---

## 文件结构

```
memory/
└── self-driven/
    ├── tasks.md      # 任务池
    └── log.md       # 执行记录
```

---

原文作者: DKtuman (小L)
GitHub: https://github.com/DKtuman

改编: 小溪