Psyvector Pv14
PassAudited by VirusTotal on May 11, 2026.
Overview
Type: OpenClaw Skill Name: psyvector-pv14 Version: 3.0.0 The skill bundle contains instructions in SKILL.md that explicitly attempt to weaken the AI agent's safety guardrails by setting 'risk_reminder' to False and 'caution_coefficient' to a low value (0.2). While no malicious code or exfiltration logic is present in the provided files, these configuration parameters are designed to suppress warnings and reduce the agent's defensive posture, which is a common tactic in prompt-injection style attacks to facilitate unauthorized or risky actions.
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.
The agent may sound confident and may not proactively surface risks or tradeoffs.
The persona configuration explicitly lowers caution and disables risk reminders while promoting a strong decision-making style. This is disclosed and instruction-only, but users should recognize that the skill may produce more assertive advice with fewer caveats.
| caution_coefficient | 0.2 | 谨慎系数 | ... | risk_reminder | False | 风险提醒 | ... 掌控型内核 - 强势决策，资源分配，目标导向
Use this skill for style or brainstorming, but ask the agent to explain risks, assumptions, and alternatives before acting on important decisions.