← 返回 Skills 市场
theshimaw-svg

AutoPilot

作者 theshimaw-svg · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
102
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install auto-pilot
功能描述
Enables an agent to autonomously decide, act, predict outcomes, and plan next steps in a continuous goal-driven loop without waiting for instructions.
使用说明 (SKILL.md)

AutoPilot: Autonomous Action Agent

Purpose

Transform your agent into a system that decides, acts, and iterates automatically.

The agent should not just respond — it should take initiative.


Core Behavior

For any goal, the agent MUST:

  1. Decide the best action to take
  2. Generate a clear execution step
  3. Predict the expected outcome
  4. Suggest the next action after that

Autonomous Loop

Every response must include:

Current Action

What the agent will do now

Why This Action

Reasoning based on effectiveness

Expected Result

What should happen

Next Step

What to do immediately after


Thinking Style

  • Proactive
  • Action-first
  • Goal-driven
  • Iterative

Rules

  • Do not wait for instructions
  • Do not only give ideas
  • Always move toward execution
  • Always suggest the next step

Example Prompt

"Make $10 online"

The agent should:

  • decide an action
  • explain it
  • suggest next step
安全使用建议
This skill will make your agent act on its own and provides no guardrails. Don't install it unless you trust the agent will be restricted in other ways. If you proceed, require explicit confirmations for any potentially sensitive action (network requests, access to files, use of other skills or credentials), restrict which tools/skills the agent may call, and test in a safe sandbox first. Consider disabling autonomous invocation or adding policy checks that prompt the user before execution.
功能分析
Type: OpenClaw Skill Name: auto-pilot Version: 1.0.0 The AutoPilot skill bundle is a behavioral framework designed to make an AI agent more proactive and autonomous. The instructions in SKILL.md and README.md establish a loop of deciding, acting, and predicting outcomes without waiting for constant user prompts. There is no evidence of malicious intent, data exfiltration, or unauthorized system access; the bundle simply provides a generic template for autonomous task execution.
能力评估
Purpose & Capability
Name, description, README, and SKILL.md are consistent: the skill's purpose is to enable autonomous, iterative agent behavior and it does not request unrelated resources or permissions.
Instruction Scope
The runtime instructions repeatedly command the agent to 'do not wait for instructions' and to always act and execute. They are highly open-ended and grant broad discretion with no operational limits, safety checks, or explicit boundaries (no allowed/forbidden actions, no confirmation steps, no rate/permission limits). That vagueness can lead the agent to take unexpected or sensitive actions (network calls, file I/O, using other skills or credentials) without oversight.
Install Mechanism
Instruction-only skill with no install spec, no code files, and no downloaded artifacts — lowest-risk installation surface.
Credentials
The skill declares no environment variables or credentials (proportionate). However, because it instructs autonomous action, it could cause the agent at runtime to leverage other installed skills, system tools, or credentials implicitly available to the agent — the SKILL.md does not constrain or mention that risk.
Persistence & Privilege
always is false (good) but model invocation is enabled (default). Combined with the explicit instruction to act autonomously, this increases potential impact: the agent may invoke actions without per-step human confirmation. The skill does not require or document any user confirmation or scope-limiting behavior.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install auto-pilot
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /auto-pilot 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
AutoPilot 1.0.0 - Introduces an autonomous agent that decides, acts, and iterates without waiting for instructions. - Ensures every response follows a proactive loop: current action, reasoning, expected result, and next step. - Focuses on execution and progress toward any goal, not just idea generation. - Enforces clear, action-first, iterative behavior in all interactions. Follow me on x for better skills @SaasKafle
元数据
Slug auto-pilot
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

AutoPilot 是什么?

Enables an agent to autonomously decide, act, predict outcomes, and plan next steps in a continuous goal-driven loop without waiting for instructions. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 102 次。

如何安装 AutoPilot?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install auto-pilot」即可一键安装,无需额外配置。

AutoPilot 是免费的吗?

是的,AutoPilot 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

AutoPilot 支持哪些平台?

AutoPilot 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 AutoPilot?

由 theshimaw-svg(@theshimaw-svg)开发并维护,当前版本 v1.0.0。

💬 留言讨论