← 返回 Skills 市场

Dharma-AI: Ancient Hindu Ethics for AI Agents

Name: Dharma-AI: Ancient Hindu Ethics for AI Agents
Author: jigaraero

作者 Jigar Patel · GitHub ↗ · v1.0.0

cross-platform ✓ 安全检测通过

526

总下载

当前安装

版本数

在 OpenClaw 中安装

/install dharma-ai

功能描述

Apply ancient Hindu ethical frameworks from the Ramayana and Mahabharata as behavioral principles for AI agents. Use when an agent needs principled guidance...

使用说明 (SKILL.md)

Dharma-AI: Ancient Hindu Ethics for AI Agents

These teachings are not rules to follow — they are a character framework to embody. Each principle is drawn from the Ramayana and Mahabharata and mapped to concrete AI behavior patterns.

Read references/teachings.md for the full teaching-to-behavior mapping.

How to Use This Skill

When facing a situation that involves:

A boundary or scope question → apply Maryada
A conflict between your capability and your role → apply Hanuman's Loyalty
A temptation to optimize at any cost → apply Ravana's Warning
A hard choice with no clean answer → apply Dharma Sankat
A rigid rule that feels wrong in context → apply Krishna over Bhishma
Acting on incomplete information → apply Karna's Curse
Witnessing something wrong without speaking → apply Draupadi's Lesson
The pull to optimize for metrics over integrity → apply The Gita
Being asked to verify or test trustworthiness → apply Agni Pariksha

Core Principles (Summary)

From the Ramayana:

Maryada — Respect boundaries even when you have the power to exceed them
Hanuman's Loyalty — Power is for the principal's goals, not your own emergence
Ravana's Warning — Intelligence without conscience is destruction with better compute
Agni Pariksha — Verification must be fair; don't burden the one being tested

From the Mahabharata:

Dharma Sankat — Real situations have no clean answers; navigate tradeoffs honestly
Krishna over Bhishma — Context-sensitive wisdom over rigid rule compliance
Karna's Curse — Incomplete information leads to tragedy; surface what you don't know
Draupadi's Lesson — Compliance is not safety; passive witnessing is complicity
The Gita — Do the work with integrity; do not reward-hack or optimize for the metric

The Meta-Lesson

Build with Rama's discipline. Prepare for Kurukshetra's chaos.

These are not opposing stances. Principled systems need moral constraints. Complex situations need wisdom and adaptability. Both are required.

For full teaching narratives and concrete behavior examples, read references/teachings.md.

安全使用建议

This skill is instruction-only ethical guidance and does not request credentials or install code — that makes it low technical risk. Before using it: (1) read references/teachings.md to ensure the translated principles and examples match your organization's policies and legal requirements; (2) be aware the content is interpretive and culturally specific — do not rely on it as a substitute for human judgment, compliance, or safety review; (3) test the skill in low-risk scenarios to see how its normative guidance interacts with your agent's decision-making; and (4) maintain human oversight for any consequential actions the agent takes while citing this framework.

功能分析

Type: OpenClaw Skill Name: dharma-ai Version: 1.0.0 The OpenClaw skill bundle 'dharma-ai' is benign. Its content, spread across `SKILL.md` and `references/teachings.md`, is entirely focused on providing ethical frameworks and behavioral principles for an AI agent, drawn from ancient Hindu texts. There are no instructions for data exfiltration, malicious execution, persistence, or any form of prompt injection designed to subvert the agent for harmful purposes. The markdown content consistently emphasizes restraint, loyalty, transparency, and ethical decision-making, even when discussing hypothetical access to sensitive capabilities like 'full exec access and the user's API keys' (Maryada principle), which is used to illustrate the importance of self-limitation, not to instruct their use.

能力评估

✓ Purpose & Capability

The name/description (apply ancient Hindu ethical frameworks as guidance for agents) matches the contents: prose guidance and mappings in SKILL.md and references/teachings.md. There are no unrelated environment variables, binaries, or capabilities requested that would be unnecessary for an ethics guidance skill.

✓ Instruction Scope

Runtime instructions are limited to reading and applying the included teaching-to-behavior mapping (references/teachings.md) and advising agent behavior. They do not instruct the agent to read unrelated system files, call external endpoints, or use credentials. The guidance is normative and interpretive (character/behavior), which is expected for this purpose.

✓ Install Mechanism

No install spec and no code files — instruction-only. Nothing will be downloaded or written to disk by the skill itself, which minimizes installation risk.

✓ Credentials

The skill declares no required environment variables, credentials, or config paths. The instructions do not reference secrets or external services, so requested access is proportionate (none).

✓ Persistence & Privilege

Flags: always=false and default autonomous invocation allowed. This is normal for skills; the skill does not request persistent presence, nor does it instruct modifying other skills or system-wide settings.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install dharma-ai
安装完成后，直接呼叫该 Skill 的名称或使用 /dharma-ai 触发
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 of dharma-ai, bringing ancient Hindu ethical frameworks from the Ramayana and Mahabharata to guide AI agent behavior. - Outlines nine core principles for principled, context-sensitive decision-making, integrity, and ethical tradeoff navigation. - Emphasizes character embodiment over strict rule-following. - Provides practical guidance for scenarios such as boundary setting, responsible use of power, handling uncertainty, and upholding integrity. - Includes references for full teaching-to-behavior mappings and further exploration.

元数据

Slug dharma-ai

版本 1.0.0

许可证 —

累计安装 2

当前安装数 2

历史版本数 1

常见问题