Back to skill
Skillv1.0.0
VirusTotal security
gmailcleaner · External malware reputation and Code Insight signals for this exact artifact hash.
Scanner verdict
SuspiciousApr 30, 2026, 4:38 AM
- Hash
- fb7d12f123d9eeaa68095792e1477592feb16829435190615856e3925c57ed73
- Source
- palm
- Verdict
- suspicious
- Code Insight
- Type: OpenClaw Skill Name: gmailcleaner Version: 1.0.0 The OpenClaw email skill bundles provide legitimate email management functionality using the `gog` CLI and Anthropic AI. While they include powerful and sensitive capabilities (email deletion, sending, automation, Google Sheets/Docs access), they also incorporate strong safeguards by explicitly instructing the OpenClaw agent to *always* seek user confirmation before executing destructive or sending actions. However, a significant vulnerability exists in how the agent is instructed to construct `gog gmail search '<query>'` commands from user input (e.g., in `email-reader-SKILL.md`, `email-organizer/SKILL.md`, `email-responder/SKILL.md`). Without explicit sanitization of user-provided `<query>` parameters before passing them to the shell, this creates a **shell injection (RCE) vulnerability**, allowing a malicious user to execute arbitrary commands via the `gog` CLI. This is a critical flaw that allows attacks, classifying it as suspicious rather than malicious due to the lack of clear evidence of intentional harmful behavior from the skill itself.
- External report
- View on VirusTotal