Back to skill
Skillv2.0.0
VirusTotal security
Feedback Learning V2 · External malware reputation and Code Insight signals for this exact artifact hash.
Scanner verdict
ReviewMay 1, 2026, 9:43 AM
- Hash
- 8ddb3554611cebfb68d25d29cfdcee1eb82ad0d86988758b983357317f2d0c75
- Source
- palm
- Verdict
- suspicious
- Code Insight
- Type: OpenClaw Skill Name: feedback-learning-v2 Version: 2.0.0 The skill implements a self-modifying feedback loop that automatically promotes user-provided text and execution errors into persistent behavioral rules ('genes') stored in JSON files. While the Python and Shell scripts (scripts/log-event.sh, scripts/analyze-patterns.py) do not contain explicit malicious payloads like exfiltration or reverse shells, the system's core logic creates a high-risk mechanism for persistent prompt injection. An attacker can influence the agent's future behavior by providing repetitive feedback that the system eventually promotes into structured instructions (genes.json) which the agent is then instructed to follow in SKILL.md.
- External report
- View on VirusTotal