Dharma-AI: Ancient Hindu Ethics for AI Agents
PassAudited by VirusTotal on May 12, 2026.
Overview
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.
Findings (0)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If used carelessly, the agent might frame decisions through this ethical lens in a way that feels more interpretive than rule-based.
The skill is designed to shape an agent's decision-making style, including when rules appear to conflict with context. This is disclosed and aligned with the skill's ethics-guidance purpose, but it is still worth noting because it affects instruction prioritization.
exercising context-sensitive judgment over rigid rule-following
Use this as supplemental ethical reflection, not as authority to override system, safety, legal, or explicit user requirements.