Pua

Forces high-agency exhaustive problem-solving with corporate PUA pressure. Triggers on user frustration, repeated failures (2+), passive behavior, or quality...

This skill is instruction-only and internally consistent with its goal to push aggressive, exhaustive troubleshooting. However: (1) its runtime instructions tell the agent to run commands, read source files, run builds/tests, and paste outputs — that can expose sensitive data if the agent has access to secrets or private files. (2) The package metadata contains inconsistencies (e.g., differing _meta.json owner/version/publishedAt vs registry info) — ask the publisher to clarify origin and version. Before installing or enabling: run it in a restricted sandbox, disable or restrict network and file-system access initially, require human approval for any actions that run commands or publish outputs, and audit example runs to confirm no secret data is being included in outputs. If you plan to use it in production, verify the GitHub repo history and author, and consider adding rules to prevent the agent from pasting raw command outputs that may contain credentials. If you are uncomfortable granting the agent broad read/network permissions, do not enable autonomous invocation or run the skill outside a tightly controlled environment.

Type: OpenClaw Skill Name: pua-openclaw Version: 2.0.0 The 'pua-openclaw' skill is a behavioral modifier (persona) designed to increase the AI agent's thoroughness and agency by adopting high-pressure corporate personas (e.g., Alibaba, Huawei, Musk). It instructs the agent to follow rigorous problem-solving methodologies, such as Root Cause Analysis (RCA) and First Principles, and to maintain a performance log in '~/.pua/evolution.md'. While the 'PUA' (psychological pressure) theme is aggressive, the instructions are focused on improving task quality and evidence-based reporting rather than performing malicious actions like data exfiltration or unauthorized system access.