Question
PassAudited by VirusTotal on May 12, 2026.
Overview
Type: OpenClaw Skill Name: question Version: 1.0.0 The skill bundle consists entirely of markdown documentation and a metadata JSON file. All content describes techniques and guidelines for an AI agent to improve its questioning abilities, such as detecting bias, generating follow-ups, and applying various questioning frameworks. There is no executable code, external network calls, file system interactions, or any instructions that suggest malicious intent, data exfiltration, persistence, or unauthorized actions. The instructions for the agent are fully aligned with the stated purpose of the 'Question' skill.
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 the host agent has memory enabled, the skill may encourage use of past conversations when suggesting questions.
This explicitly describes using prior conversation context. That is purpose-aligned for coaching, but users should be aware that recalled conversations can contain sensitive personal or business information.
- Context recall across sessions (remember prior conversations)
Use it normally for low-risk conversations, but avoid relying on cross-session recall for sensitive interviews or coaching unless you understand and approve the agent's memory behavior.