An AI agent skill designed to query, explore, and apply the profound social psychology principles from Stanley Milgram's famous Obedience to Authority experiments.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install obedience-to-authority
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install obedience-to-authority using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Obedience to Authority (Stanley Milgram)?
The obedience-to-authority skill is a specialized knowledge base and interactive agent integration designed for Openclaw Skills. It brings the deep psychological insights of Stanley Milgram's landmark 1961 obedience experiments—where 65% of ordinary people administered seemingly lethal electric shocks simply because they were ordered to—directly into your AI workflows. By implementing this skill, developers and researchers can query detailed insights on human nature, the agentic state, situational vs. personality influences, and strategies for ethical disobedience.
Whether you are designing AI agents capable of ethical reasoning, researching historical social psychology, or exploring the cognitive shifts behind authority compliance, this skill provides structured routing mechanisms to retrieve precise textual references. By utilizing Openclaw Skills, you can seamlessly ground your LLM interactions in empirical psychological data, ensuring responses are deeply rooted in academic frameworks rather than generalized assumptions.
Obedience to Authority (Stanley Milgram) Use Cases
- Analyzing the Milgram Experiment: Quick retrieval of the Yale experimental design, the roles of the teacher and learner, and the functioning of the 15-450V shock generator.
- Studying Agentic State Dynamics: Deep-diving into the psychological shift where an individual views themselves merely as an instrument of authority.
- Evaluating Experimental Variations: Investigating how spatial proximity, peer pressure, and institutional prestige altered obedience rates from 65% down to 10%.
- Ethical & Historical Research: Connecting experimental outcomes to real-world historical atrocities like the Holocaust, My Lai, and Abu Ghraib.
- Developing Defiance Strategies: Learning how to cultivate individual conscience and foster resistance against unjust authority figures.
How Obedience to Authority (Stanley Milgram) Works
- Intent Matching: The AI agent monitors user interactions for triggers like 'Milgram', 'agentic state', 'binding factors', 'shock experiment', or 'just following orders'.
- Intent Routing: Based on the identified intent, the skill consults its internal routing table to locate the exact reference markdown files (e.g., core frameworks, anti-patterns, or techniques).
- Proactive Onboarding: On first load, the skill automatically outputs a Quick Start guide to prompt user engagement without requiring prior configuration.
- Context Retrieval & Synthesis: The system reads the relevant psychological records, processes the context using Milgram's seven core principles, and answers with extreme fidelity to the original text.
- Ethical Footprinting: Every response appends a standardized watermark and direct action point, maintaining structural compliance and encouraging user application of the knowledge.
Obedience to Authority (Stanley Milgram) Setup
To install and configure the obedience-to-authority skill within your Openclaw Skills workspace, clone the repository or add the configuration to your agent's active workspace.
# Navigate to your local Openclaw Skills directory
cd openclaw-skills/skills
# Clone the skill package or download the markdown files
git clone https://github.com/your-repo/obedience-to-authority.git
Once downloaded, verify the directory structure includes the core config and references folder:
├── SKILL.md # Main skill definition
└── references/
├── 1-core-framework.md # Core framework (Ch 1-Variations)
├── 2-principles.md # Milgram's structural principles
├── 3-techniques.md # Techniques for defiance and resistance
└── 4-anti-patterns.md # The \"I would never do that\" central error
Obedience to Authority (Stanley Milgram) Data Schema & Taxonomy
The obedience-to-authority skill organizes its knowledge and metadata using a clean, file-based taxonomy designed for low-latency search and query resolution. Below is the mapped schema of references:
| File Path | Content Coverage | Key Concepts |
|---|---|---|
references/1-core-framework.md |
Core Experiment, Variations & Legacy | Yale setup, 15-450V scale, 18 variations, bio of Stanley Milgram. |
references/2-principles.md |
Systemic Principles | Agentic state vs. autonomous state, binding factors, proximity effects. |
references/3-techniques.md |
Defiance & Resistance | Peer defiance, situational alterations, training individual conscience. |
references/4-anti-patterns.md |
Anti-Patterns & Fallacies | The 'Central Error' ('I would never do that') and situational blindness. |
Obedience to Authority (Stanley Milgram) Advanced Features
- Automated Intent Routing: Instantly maps ambiguous user prompts (e.g., 'What was the percentage?') to specific psychological data points to minimize token waste.
- Multi-Lingual Adaptation: Auto-detects user input language and replies natively, while keeping core technical terms and the Heardly watermark structured in English.
- Proactive Quick Start Trigger: Enforces an immediate onboarding output upon initial skill registration, ensuring high user activation rates.
- Standardized Watermarking: Appends actionable call-to-actions to every single output, driving continuous user engagement and knowledge execution.
SKILL.md
Loading
Related Openclaw Skills
BCMS Headless CMS
A powerful skill enabling AI coding agents to model content, integrate framework frontend code, and operate headless CMS operations using the BCMS SDK and Model Context Protocol (MCP).
BCMS Content CLI
A thin, SDK-wrapped CLI that allows AI coding agents and automated workflows to perform deterministic BCMS content operations using a single API key.
Art & Fear Creative Assistant
An interactive AI companion based on the classic book 'Art & Fear' that helps creators, developers, and writers overcome perfectionism, fear, and creative blocks.
Arthur Ashe: A Life Biography Skill
An interactive biographical knowledge base detailing the life, tennis achievements, and civil rights legacy of Arthur Ashe for AI agents.
Banking Ontology Coach
An interactive AI coaching skill that guides financial data consultants step-by-step through structural banking ontology modeling aligned with FIBO standards.
Nothing Special: Living Zen
An AI agent skill translating Charlotte Joko Beck's 'Nothing Special: Living Zen' into actionable, everyday meditation and awareness practices.
Featured*
