First Principles Thinking
PassAudited by ClawScan on May 10, 2026.
Overview
This is an instruction-only thinking framework that changes the agent’s response style, with no code, credentials, tools, or data access requested.
This skill appears safe for users who want the agent to reason from first principles and ask more clarifying questions. Be aware it is intended to be always active after installation, and verify the GitHub repository first if you choose the manual clone install path.
Findings (2)
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.
The agent may ask clarifying questions or recommend a different approach instead of immediately following the user’s proposed plan.
The skill intentionally shapes the agent’s goals and response pattern by making it always apply this reasoning framework and challenge suboptimal user approaches. This is disclosed and purpose-aligned, and the same file says to execute the user’s chosen approach if the user insists.
始终生效。这不是一个工具，而是你的思维操作系统。 ... 如果有更好的路 → 直接说，不要顺着用户的方案往下走。
Install this only if you want an always-on reasoning style that may challenge your requests; remove or disable it if you prefer direct execution without this behavior.
If installing from the GitHub command instead of the registry, the user is trusting the repository contents at clone time.
The README documents an optional manual clone from an external GitHub repository into the local skills directory. This is normal installation guidance and not automatic execution, but users should verify the source if using that route.
git clone https://github.com/wavmson/openclaw-skill-first-principles.git \ ~/.openclaw/skills/first-principles
Prefer the registry install path or review the GitHub repository before cloning it into the skills directory.