Back to skill
Skillv0.1.0
VirusTotal security
Aegis Shield · External malware reputation and Code Insight signals for this exact artifact hash.
Scanner verdict
BenignApr 29, 2026, 3:48 AM
- Hash
- 5e511424d89fce1ff875f65b6754b32f0eb46c3314aa1c270663c985dc0ddab6
- Source
- palm
- Verdict
- benign
- Code Insight
- Type: OpenClaw Skill Name: aegis-shield Version: 0.1.0 The OpenClaw AgentSkills bundle 'aegis-shield' is designed as a security tool to screen untrusted text for prompt injection, data exfiltration, and tool abuse patterns before writing to memory. The `SKILL.md` documentation clearly outlines its security purpose and provides safe usage guidelines. The `openclaw-safe-memory-append.js` script implements this by performing linting (checking for imperative language, dangerous tool commands, secret-like keywords, and authority laundering attempts) and scanning (via a local library) of input text. It then either quarantines or sanitizes and appends the text to specific, controlled memory files within the OpenClaw workspace. All file system operations are confined to the intended memory directories, and there is no evidence of malicious intent such as unauthorized data exfiltration, arbitrary command execution, or persistence mechanisms. The script's functionality is entirely aligned with its stated purpose of enhancing agent security.
- External report
- View on VirusTotal