Back to skill
Skillv1.0.0
VirusTotal security
SELF LEARNING SKILL V2 · External malware reputation and Code Insight signals for this exact artifact hash.
Scanner verdict
SuspiciousApr 30, 2026, 6:23 AM
- Hash
- 0b201ce53e9a8c2abcddf92f7c44cf1b6aa9f1ac24fc6bb0b0af30aeae0cd8f1
- Source
- palm
- Verdict
- suspicious
- Code Insight
- Type: OpenClaw Skill Name: self-learning-skill-v2 Version: 1.0.0 The skill bundle implements a meta-learning framework for an AI agent to track errors, perform root cause analysis, and update its own capability matrix. It is classified as suspicious because SKILL.md and ERROR_LOG.md explicitly instruct the agent to "search for tokens in multiple locations" (建立 token 搜索机制，多位置查找) to resolve authentication failures. While this is plausibly intended to fix GitHub publishing issues, the broad instruction to hunt for credentials across the filesystem is a high-risk behavior. Additionally, EXECUTE.md defines scheduled automation tasks (e.g., daily checks at 22:00) that imply persistent background activity.
- External report
- View on VirusTotal