Back to skill
Skillv2.0.4
VirusTotal security
Checkmate · External malware reputation and Code Insight signals for this exact artifact hash.
Scanner verdict
SuspiciousApr 30, 2026, 4:21 AM
- Hash
- cca4969a0a82e0ddd5347d59f7fe57727ceb689284450bb7d978adfcb70495d2
- Source
- palm
- Verdict
- suspicious
- Code Insight
- Type: OpenClaw Skill Name: checkmate Version: 2.0.4 The OpenClaw 'checkmate' skill is classified as suspicious due to a significant prompt injection vulnerability that could lead to Remote Code Execution (RCE). The skill is explicitly declared as 'high-privilege' in SKILL.md and README.md, granting spawned worker agents full host-agent runtime, including `exec` (arbitrary shell commands). User-provided input (initial task, task edits, and iteration feedback) is directly incorporated into the prompts for these worker agents (e.g., `{{TASK}}`, `{{FEEDBACK}}` in `prompts/worker.md`). A malicious user could craft input that, when injected into the agent's prompt, causes the agent to execute arbitrary commands on the host system. While the skill's documentation transparently warns about these risks and the `scripts/run.py` orchestrator uses `subprocess.run` safely for its own CLI calls, the design inherently allows for RCE via prompt injection against the agent, making it a critical vulnerability rather than intentional malware.
- External report
- View on VirusTotal