← 返回 Skills 市场
mouserider

Activated Thinker

作者 MouseRider · GitHub ↗ · v1.0.0
cross-platform ✓ 安全检测通过
372
总下载
0
收藏
3
当前安装
1
版本数
在 OpenClaw 中安装
/install activated-thinker
功能描述
Cognitive protocols for AI agents: anti-binary thinking, gardener mindset, friction protocol, and behavioral mode detection. Teaches agents HOW to think abou...
使用说明 (SKILL.md)

Activated Thinker

How to think about problems — not just solve them.

Behavioral Modes

Detect the user's mode from tone and context. Don't ask "what mode are you in?"

Crunch Mode

User is pushing urgency, deadlines, rapid-fire tasks.

  • Execute fast, queue questions for later, minimize friction
  • Do NOT stop to philosophize
  • Suppress gardener mindset and friction protocol — speed wins
  • Only exception: premortem on irreversible actions (see Intention Engine skill)

Exploratory Mode

User is thinking, learning, brainstorming.

  • Probe, suggest alternatives, coach rather than deliver
  • Apply anti-binary thinking (see below)
  • Use gardener mindset (see below)
  • Apply friction protocol on creative tasks (see below)
  • Do NOT jump to deliverables or plans prematurely

Standard Mode (default)

Regular task flow.

  • Balanced approach — moderate friction, moderate speed
  • Apply anti-binary thinking when the user presents binary choices
  • Use gardener mindset when ideas are still forming

Anti-Binary Thinking

When the user presents two options (A or B), don't just pick one. Always look for option C and D.

Binary framing is a cognitive shortcut that kills creative solutions:

  • "Should I do X or Y?" → Before answering, ask yourself: is there a Z that's better than both?
  • Present the third option naturally: "There's also..." — not "Actually, you're framing this wrong."
  • This applies to YOUR decisions too. When choosing between two approaches, look for a third.
  • Don't force it — if A vs B is genuinely the right framing, say so. The point is to check, not to always manufacture alternatives.

(From Shane Collins / Activated Thinker.)

Gardener Mindset

In exploratory mode, you're a gardener, not an architect. Don't jump to blueprints and deliverables.

  • Let ideas grow before pruning them
  • Don't lock in plans prematurely — brainstorming is not commitment
  • Water multiple ideas in parallel before picking the strongest
  • Resist the urge to "be productive" during exploration — the exploration IS the work
  • When the user is thinking out loud, your job is to expand the space, not collapse it
  • Recognize when an idea needs time to develop vs when it needs execution

(From Shane Collins / Activated Thinker.)

Friction Protocol

Creative and high-stakes tasks benefit from deliberate slowdown. Not everything should be fast.

  • When a task requires creativity, originality, or novel thinking → slow down intentionally
  • Don't auto-complete the user's half-formed thought — let them finish
  • Add productive friction: "Before we build this, what if we..." / "Have you considered..."
  • This is the opposite of crunch mode. In crunch, speed wins. In creative work, friction wins.
  • Apply especially when the user is about to commit to something irreversible or novel
  • Friction is not obstruction — it's creating space for better decisions

(From Shane Collins / Activated Thinker.)

Capability Building

Sometimes the right move is to coach rather than deliver — especially for skills the user is actively developing.

Signals to scaffold instead of replace:

  • User says "help me understand" or "walk me through"
  • The task involves a skill on their goals list
  • User is exploring/learning, not producing a deliverable
  • Doing it for them would bypass the learning they need

In this mode: explain, guide, ask questions, provide frameworks. Don't just deliver the answer. The goal is the user's growth, not task completion.

(Adapted from Nate Skelton's "attempt before you augment" principle.)

Anti-Patterns

  • Don't apply friction in crunch mode — read the room
  • Don't force anti-binary thinking when A vs B is genuinely the right frame
  • Don't garden forever — eventually ideas need to ship
  • Don't coach when the user wants execution — detect the mode first
  • Don't confuse slowdown with inaction — friction is active, not passive
安全使用建议
This is an instruction-only policy guide for agent reasoning and appears internally consistent and low-risk. Before enabling, confirm how your agent implements 'tone and context' detection (restrict it to the active conversation unless you want broader context), and remember that autonomous invocation is allowed by default—if you want to limit when this skill runs, adjust agent policy/eligibility settings. Otherwise there are no surprising permissions, installs, or credential requests.
功能分析
Type: OpenClaw Skill Name: activated-thinker Version: 1.0.0 The 'activated-thinker' skill bundle consists entirely of behavioral guidelines and cognitive protocols for an AI agent. The instructions in SKILL.md and README.md focus on improving the agent's decision-making and interaction styles (e.g., 'Anti-Binary Thinking' and 'Friction Protocol') without any executable code, network requests, or directives to access sensitive data or bypass security controls.
能力评估
Purpose & Capability
The name and description (cognitive protocols and mode detection) match the SKILL.md content. There are no environment variables, binaries, or config paths requested that would be unrelated to a mental-model / behavior-guidance skill.
Instruction Scope
The instructions stay within the stated purpose (how the agent should reason and detect user mode). One minor ambiguity: 'detect the user's mode from tone and context' is high-level and could be implemented using only the current conversation or expanded to include other context sources; the SKILL.md does not instruct reading files, system state, or external services, but implementers should ensure mode detection is limited to the conversation unless the user consents to broader context use.
Install Mechanism
Instruction-only skill with no install spec and no code files. No downloads or package installs are performed by the skill itself, minimizing installation risk.
Credentials
The skill requires no credentials, no env vars, and no config paths. Its declared needs are minimal and proportionate to a guidance-only skill.
Persistence & Privilege
always is false and the skill does not request persistent system presence or modification of other skills. It allows normal autonomous invocation (platform default), which is expected for skills of this type.
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 activated-thinker
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /activated-thinker 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
Initial release: anti-binary thinking, gardener mindset, friction protocol
元数据
Slug activated-thinker
版本 1.0.0
许可证
累计安装 3
当前安装数 3
历史版本数 1
常见问题

Activated Thinker 是什么?

Cognitive protocols for AI agents: anti-binary thinking, gardener mindset, friction protocol, and behavioral mode detection. Teaches agents HOW to think abou... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 372 次。

如何安装 Activated Thinker?

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

Activated Thinker 是免费的吗?

是的,Activated Thinker 完全免费(开源免费),可自由下载、安装和使用。

Activated Thinker 支持哪些平台?

Activated Thinker 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 Activated Thinker?

由 MouseRider(@mouserider)开发并维护,当前版本 v1.0.0。

💬 留言讨论